WEBVTT

1
00:00:00.000 --> 00:00:01.000
It's it's it's

2
00:00:01.180 --> 00:00:02.710
Jim Donlon | NSF: it fills up a bit.

3
00:00:03.900 --> 00:00:04.980
Jim Donlon | NSF: Pigs,

4
00:00:26.000 --> 00:00:32.760
Jim Donlon | NSF: hey? Welcome, everybody! We're just going to give the the room a chance to fill up as people join coming out of the waiting room

5
00:00:35.940 --> 00:00:53.069
Jim Donlon | NSF: while people are joining. I'll just call your attention to the procedure we're going to use for Q. And A. I'll mention this once. We're well up to, but just give you an idea. So as always, if you're wondering about the the program and Haven't yet had time to see the the solicitation. You'll find

6
00:00:53.080 --> 00:00:57.440
Jim Donlon | NSF: everything that we've published so far on the program page. Here

7
00:00:58.950 --> 00:01:01.339
Jim Donlon | NSF: we will be posting the slides

8
00:01:01.660 --> 00:01:11.450
Jim Donlon | NSF: and the recording of the Webinar for everybody to reference. Later on, if you have colleagues who are unable to attend, or if you want to revisit the Webinar,

9
00:01:12.620 --> 00:01:16.220
Jim Donlon | NSF: the room still filling up, would just give up just a few more beats.

10
00:01:34.920 --> 00:02:01.940
Jim Donlon | NSF: Good afternoon, everybody, and thank you for joining. I'm Jim Donlan, program director and Nsf. Information and Intelligence Systems, and the program Director for the National Ai Research Institute's program on behalf of the entire working group. I'm: pleased to present you this webinar. The Webinar is scheduled to begin now at one thirty, and we're going to end it between two hundred and thirty and three, depending on how long. It takes us to exhaust all questions. We're going to present for approximately fifty minutes, and then we'll begin taking questions for the remainder of the time.

11
00:02:01.950 --> 00:02:31.640
Jim Donlon | NSF: So here you can see the procedure for Q. And A. You may submit your questions through the Q. And A. Module. So if you've never been on a zoom webinar you should find on your little toolbar there a button, as illustrated, called Q. A. Um. There's place there where you can type your questions in, and you'll also be able to see and even upload, if you like, the questions of other people. So we're going to be answering the questions that accumulate in there. As many of them live as we can. Some of them will answer directly in the module, if they're sort of straight,

12
00:02:31.650 --> 00:02:46.660
Jim Donlon | NSF: so we're going to answer as many program-wide questions as we can. If your question is about a particular proposing scenario we're likely to defer that to one on one contact between you. And the program leads one hundred and one.

13
00:02:52.560 --> 00:02:54.110
Jim Donlon | NSF: There we go.

14
00:02:54.120 --> 00:02:55.230
Jim Donlon | NSF: Two

15
00:02:55.240 --> 00:03:23.230
Jim Donlon | NSF: Ai research Institutes is a multi-section program represented by program directors across Nsf. As well as by program contacts representing the artificial prior artificial intelligence priorities of our partners. Several program contacts are attending the Webinar today. You're going to hear from the contacts listed here. These are the principal coordinators of their respective themes. We're also joined by representatives from partner organizations in the program solicitation. You'll find a full list of program contacts for the themes and their organization affiliations

16
00:03:23.240 --> 00:03:42.190
Jim Donlon | NSF: before we get into all that, i'm very honored to uh to uh introduce Dr. Catherine ponch ponchen often. Uh Dana Sep directors. Uh joined us to give you a welcome and and some encouragement about the program. So for that i'm going to drop the slides, and uh and hand it over to Dr. Bonf, sir.

17
00:03:42.480 --> 00:03:46.390
Sethuraman Panchanathan: Thank you, Jim, and it's truly a pleasure to be with all of you today,

18
00:03:46.400 --> 00:03:52.290
Sethuraman Panchanathan: and welcome to Nsf. And welcome to this Webinar. This is a very, very important

19
00:03:52.300 --> 00:04:10.469
Sethuraman Panchanathan: Webinar, as you can imagine, because Ai, as you all know very well already is a very important hot topic right now, and this is something that I take a lot of pride in talking about. Nsf. Which has made what is possible today.

20
00:04:10.480 --> 00:04:23.919
Sethuraman Panchanathan: We have investments have been consistent over the last several decades, and what you hear and see about Ai today is because of the investments that Nsf has made over the last several decades.

21
00:04:23.930 --> 00:04:31.289
Sethuraman Panchanathan: I mean, if you look at Nsfs investments in computing eighty percent of the non-defense related research in computing

22
00:04:31.300 --> 00:04:50.140
Sethuraman Panchanathan: and Ai is made by Nsf. And that has made possible significant discoveries, innovations, outcomes, and many of you, i'm sure have been part of that, and we're excited to have you here to again talk to you about the Ai Institute's program when a kid Nsf: one

23
00:04:50.150 --> 00:04:57.389
Sethuraman Panchanathan: uh this, about three years ago. Now you know one of the you know pioneering programs that are being thought about to be launched,

24
00:04:57.400 --> 00:05:07.309
Sethuraman Panchanathan: what the Ai institutes, and I said that Ai Institute should be configured such that it is something that impacts every part of our nation,

25
00:05:07.320 --> 00:05:22.100
Sethuraman Panchanathan: and that makes possibilities to unleash talent and ideas which all of you agree is democratizing is everywhere. How do we unleash that possibility through this vehicle of Ai institutes through public private partnerships,

26
00:05:22.130 --> 00:05:30.789
Sethuraman Panchanathan: and as well as context, that are embedded in each of these locations that the Ai institutes are, so that we might unleash those innovations,

27
00:05:30.800 --> 00:05:46.089
Sethuraman Panchanathan: not only in terms of the Col. Ai activities, but also Ai related to many of the disciplinary innovations and as well as solving problems, grand challenge problems for society and humanity and national security. And of course, the economy.

28
00:05:46.100 --> 00:06:04.789
Sethuraman Panchanathan: Now you look at the first. Ah, seven Ai Institute that launched um, You know I had ah employed folks to make sure that you know we see partnerships while, and Nsf. Is investing close to one hundred million dollars. Ah, you know, to tune of twenty million dollars and institute. You know five of them that Nsf. Had funding, for

29
00:06:04.800 --> 00:06:07.189
I wanted to make sure that it was not limited by that.

30
00:06:07.200 --> 00:06:17.479
Sethuraman Panchanathan: We want to make sure that we are not with other entities to be able to bring resources, and therefore expand the scale of what Ai Institutes can do; and in the first situation we are two of them

31
00:06:17.490 --> 00:06:35.470
Sethuraman Panchanathan: fully funded by the Us. Department I'm. A country in addition to other partnerships we had with the other Ai Institutes, and the seven of them were the first tranche of the Ai Institute, and in the second cycle we were able to bring six additional institutes. In addition to the Five Nsf. Was going to invest in a total of eleven,

32
00:06:35.480 --> 00:06:56.040
Sethuraman Panchanathan: and again, two more of that was Usda funded. In fact, one of them was made possible through an investment that was made by a collection of industry, a consortium of industry, intellectual and Amazon, bringing together that scale of investment of twenty million dollars. So, as you can see through partnerships with the part of transportation.

33
00:06:56.200 --> 00:07:03.789
Sethuraman Panchanathan: You know a Nasa and and I age and Deal, e and and Nist, and no one others

34
00:07:03.800 --> 00:07:09.790
Sethuraman Panchanathan: clear this room for us to do some amazing things as we think about Ai Institutes and what it can do for the future.

35
00:07:09.980 --> 00:07:19.489
Sethuraman Panchanathan: Thanks to all of you and your work, and your interest is why we will be able to ensure success of this program as they continue into the future.

36
00:07:19.500 --> 00:07:21.190
Sethuraman Panchanathan: The third iteration again,

37
00:07:21.200 --> 00:07:39.829
Sethuraman Panchanathan: we had one of the seven institutes to probably now they're keeping count seven plus eleven, plus seven, the twenty-five institutes that i'm, pleased to say spans almost forty states or plus forty prostates, or even more, and then in Washington, Dc. One

38
00:07:39.840 --> 00:08:00.619
Sethuraman Panchanathan: and poured on others. So i'm so excited by the fact that the reach of this clearly meets the original mission and goals of what we want to accomplish is that it touches many parts of our great nation, and unleashes possibilities everywhere, but also takes advantage of the context of the Ai expertise that resides all across our nation.

39
00:08:00.630 --> 00:08:30.179
Sethuraman Panchanathan: Now we are here launching the next iteration, and again, the partnerships are tremendous. When you look at the new solicitation that you're building for my initial network of Ai Institutes, we are focused on three exciting new areas where Ai will be a powerful tool for progress. One of them is with assignments foundation, where we can bring the Ai in its approach to analyzing the immense volume of high quality astronomical data coming from major astronomical,

40
00:08:30.190 --> 00:08:37.329
Sethuraman Panchanathan: you know. Not too long ago I was there in Hawaii, in Maui, launching the deepest that a sword telescope.

41
00:08:37.340 --> 00:08:58.289
Sethuraman Panchanathan: And when you look at these telescopes and look at the amount of data that comes in, there is so much possibilities there to be able to use Ai tools and techniques to be able to achieve the kinds of discoveries that is made possible because of the fact that you're able to bring in these advanced tools to analyzing the big data

42
00:08:58.480 --> 00:09:11.280
Sethuraman Panchanathan: that comes from these telescopes, whether it's in Chile or Hawaii, and and other other locations. So even achievements like the image of the Black Hole just a few years ago, or the recent Nano Grab announcement.

43
00:09:11.290 --> 00:09:27.060
Sethuraman Panchanathan: You know that that that we that we sort of brought out a few weeks ago to check gravitational waves. This is a great time to be an astronomer, and the new Ai tools are going to power, exciting new discoveries. Now, likewise, we are partnering with intel corporation

44
00:09:27.210 --> 00:09:41.689
Sethuraman Panchanathan: to personal new discoveries in materials, research powered by Ai. This is a very exciting area of research and new materials with special properties, are going to be the building blocks of amazing new products and technological capabilities.

45
00:09:42.010 --> 00:10:04.050
Sethuraman Panchanathan: We are also proud to partner with capital one for the very first time to tackle what are shooting Today's most pressing research issues for the future of Ai. How to improve the fundamental design and understanding of rapidly advancing Ai models, so that we can guide and prioritize development of Ai capabilities that reflect the true understanding of the world,

46
00:10:04.220 --> 00:10:10.810
Sethuraman Panchanathan: respond to human instruction and remain aligned with human and societal values. The

47
00:10:10.950 --> 00:10:16.610
Sethuraman Panchanathan: in July I participate in a briefing for senators on the state of Ai innovation.

48
00:10:16.700 --> 00:10:22.090
Sethuraman Panchanathan: We have a fantastic bipartisan engagement from from Congress on this issue.

49
00:10:22.100 --> 00:10:26.279
Sethuraman Panchanathan: As I said in the beginning, this is a critical time for Ai research.

50
00:10:26.600 --> 00:10:36.309
Sethuraman Panchanathan: This has always been an important area of research innovation. We all know that you're all part of it. You're all making possible amazing things, and you're going to make possible even greater things.

51
00:10:36.540 --> 00:10:48.840
Sethuraman Panchanathan: So now we are seeing new attention from the public and from lawmakers who want to know more about what Ai needs for how we work for our economy and for our national competitiveness.

52
00:10:48.930 --> 00:11:04.870
Sethuraman Panchanathan: So that is why this national imperative to advance Ai is such a development, and the Ai Institutes are a vital part of how we fulfill that mission and carry out Nss promise to advance science and technology for the benefit of society.

53
00:11:05.360 --> 00:11:24.559
Sethuraman Panchanathan: So I will tell you categorically that these institutes build on a rich foundation of us leadership in Ai research innovation. They are showing how important it is to build large collaborative communities to advance Ai on how important convergent research is to solving major challenges.

54
00:11:24.570 --> 00:11:36.680
Sethuraman Panchanathan: So all these institutes, we know, are essential to how we prepare students at the end of the day. It's about talent and ideally, and we at full force and full scale all across our nation.

55
00:11:36.710 --> 00:11:44.089
Sethuraman Panchanathan: How do we educate your future workforce and strengthen our nation's capacity for ai powered productivity. The

56
00:11:44.100 --> 00:11:48.709
Sethuraman Panchanathan: so i'm very excited to be here with all of you. I want to thank the colleagues

57
00:11:48.860 --> 00:11:49.990
Sethuraman Panchanathan: from any cell

58
00:11:50.000 --> 00:12:07.549
Sethuraman Panchanathan: the amazing book that is being done by Jim Donnelly, Michael Littman, Andreas, Berlin, Hector Munoz, Avalia, and Saddam Hussein, who's working really hard to make sure that they are providing you those opportunities that surely is going to transform

59
00:12:07.570 --> 00:12:09.540
Sethuraman Panchanathan: the discipline of Ai

60
00:12:09.900 --> 00:12:13.560
Sethuraman Panchanathan: and the impact of Ai in every aspect.

61
00:12:13.650 --> 00:12:24.100
Sethuraman Panchanathan: Our daily lives. So thank you for joining us. Thank you for the fact that you are part of this potential for transformational progress in Ai research.

62
00:12:24.110 --> 00:12:39.230
Sethuraman Panchanathan: The Ai Institutes are the bedrock of a prosperous future, enabled by Ai capabilities that you are envisioning, designing, developing, and implementing. And we cannot wait to see the creative partnerships you build to make all that possible.

63
00:12:40.190 --> 00:12:53.260
Sethuraman Panchanathan: Michael and Um and Andreas, Hector Sedar and Jim thank you for the opportunity for being with all of you today, and I let you do the real work and and get out of the way. So thank you all,

64
00:12:53.270 --> 00:13:00.770
Jim Donlon | NSF: Dr. Punch, and often thank you for your encouragement and your leadership. We really appreciate the environment that makes us possible. Thanks so much for joining us.

65
00:13:01.080 --> 00:13:31.019
Jim Donlon | NSF: So to start the webinar I'd like to introduce. Now Dr. Michael Litman, who's the division director for the division of Intelligent and information systems in slice. That's the Home Division for the program. Um, and he's going to convey some observations on Ai strategy, and where the program fits in Michael is going to be representing not only this the home division for the program, but also representing Dr. Margaret Martin Nosey. She's our Nsf. Assistant director, Precise. Unfortunately, Margaret was called away just the last few minutes, and Michael

66
00:13:31.030 --> 00:13:40.049
Jim Donlon | NSF: gladly agreed to step in and present some remarks on Margaret's behalf. So with that i'm gonna pass it off to Michael, sir.

67
00:13:40.290 --> 00:13:51.790
Michael Littman: Yeah, thanks so much, Jim. So just to be clear. The remarks were prepared for uh, Margaret, and and by Margaret, i'm going to be trying to fill in at the last moment. But I will add a little bit of my own kind of

68
00:13:51.800 --> 00:14:10.350
Michael Littman: angle on some of this as well, so I am really happy to join and welcome you, not only representing the A. D. For the program's Home Director and the Assistant Director, but on behalf of all the other assistant directors of all the different directorates in Nsf, each of which is a partner in this nsf-wide initiative.

69
00:14:10.460 --> 00:14:17.729
Michael Littman: So, as a quick aside I personally am delighted to be part of this process as a lifelong Ai researcher at a year-long Nsf:

70
00:14:17.810 --> 00:14:27.629
Michael Littman: This solicitation is actually the first i've gotten to be a part of at Nsf. And I'm. Very excited by the themes and by the opportunities, and to see what everyone's going to do with it.

71
00:14:28.160 --> 00:14:29.290
Michael Littman: All right.

72
00:14:29.300 --> 00:14:50.719
Michael Littman: So as Dr. Punch Nathan emphasized, this program is a joint effort between the National Science Foundation partners in the interagency and the research community to establish Ai institutes that serve as national nexus points for collaborative ai efforts, spanning institutions of higher education, federal agencies, industry and nonprofits and foundations.

73
00:14:50.790 --> 00:15:02.140
Michael Littman: The program and the institutes funded by it, aligned in some significant way with each of the objectives in the National Ai R. And D Strategic plan that originally inspired the Ai Institutes

74
00:15:04.090 --> 00:15:14.979
Michael Littman: i'll re-emphasize that we're immensely proud of the twenty five institutes funded to date in this program over the previous three rounds, and continuing with this new funding opportunity,

75
00:15:14.990 --> 00:15:27.940
Michael Littman: each institute is a multidisciplinary multi-stakeholder team focused on advancing our understanding of Ai and developing the future Ai workforce. While addressing some of society's grand challenges.

76
00:15:29.540 --> 00:15:37.229
Michael Littman: In this new solicitation, you will see that there are that the funding opportunities span new institutes to be created over the coming two years

77
00:15:37.880 --> 00:15:55.669
Michael Littman: First and Nsf. Is partnering with the Simon's foundation with a goal to fund two institutes in two thousand and twenty four that advance ai and the astronomical sciences together to unlock new discoveries from the high volume of high-quality and multi-spectral data that is coming from major astronomical facilities.

78
00:15:56.280 --> 00:16:06.079
Michael Littman: Second Intel Corporation has joined and nsf to fund an institute in two thousand and twenty five focused on Ai advances and discovery in materials research.

79
00:16:06.090 --> 00:16:20.470
Michael Littman: Today's materials, research opportunities can push our quest for the advancement of Ai knowledge and methods for scientific discovery, while Ai has great potential to help us discover the building blocks for beneficial new products and technological capabilities,

80
00:16:20.870 --> 00:16:37.520
Michael Littman: and in strengthening Ai, we're pleased to partner with capital one and potentially with one or more federal agencies to fund two, or maybe even more institutes in two thousand and twenty five to tackle. What will certainly be a continuing and long-range project in Ai research.

81
00:16:37.530 --> 00:16:50.010
Michael Littman: And that is, how do we improve the fundamental design and understanding of rapidly advancing Ai models, so that we guide and prioritize development of Ai capabilities in the most desirable directions.

82
00:16:50.030 --> 00:16:58.989
Michael Littman: So this is going to require ai that is grounded in a true understanding of the world, meaning that it understands the concepts that it reasons over

83
00:16:59.000 --> 00:17:14.410
Michael Littman: it's instructible in that it responds appropriately to human feedback or instruction, and that we can ensure we'll operate consistently with human values and societal expectations, or what is referred to in the ai community as the alignment imperative.

84
00:17:14.800 --> 00:17:22.640
Michael Littman: This theme is explicitly aligned with the ever-present imperative that we attain high levels of trust in the technologies that we develop

85
00:17:22.740 --> 00:17:29.090
Michael Littman: the theme aspires to trust via these principles of grounding instructibility and alignment with our values.

86
00:17:29.270 --> 00:17:42.849
Michael Littman: Such strong and robust Ai can be better made to implement more effective and trustworthy assistance to humans. For example, in instruction, tutoring and training, or in explaining their understanding or recommendations as quoted from the solicitation.

87
00:17:43.430 --> 00:17:48.119
You'll hear all about these themes in greater detail from the program contacts.

88
00:17:48.550 --> 00:17:53.210
Michael Littman: So i'll hand the microphone back to the program team. But I would like to leave you with this, thought.

89
00:17:53.470 --> 00:18:09.019
Michael Littman: I'm really proud of how our Ai Institutes and other federally. Funded research projects lead in laying the groundwork for a government priority in trustworthy and beneficial Ai. So we hope to continue to see the effects continue to grow as the program expands.

90
00:18:09.100 --> 00:18:18.959
Michael Littman: Thank you to the whole program team contacts and thank you to our partners for your work towards this ideal. And of course, thank all of you for joining us today.

91
00:18:20.140 --> 00:18:39.509
Jim Donlon | NSF: Thank you, Michael, for your leadership as well. Making this a program possible in Iis. This Webinar is intended to familiarize participants with the new solicitation in the program, and to help you prepare to submit proposals that are consistent with the goals of the program. We'll continue with the overview and structure of the program as in the sequence. You see here.

92
00:18:39.640 --> 00:18:56.739
Jim Donlon | NSF: Um, one of the things I want to. Um. I want to emphasize here is that you should be referencing both the solicitation and the program. I'm sorry the Nsf. Proposal and award policies and procedures can either the Phdpg as you consider submitting to the program.

93
00:18:56.750 --> 00:19:07.939
Jim Donlon | NSF: Once we get through all this material, then the remaining time is going to be spent in Q. And A. As we cover it up front. So as we as we begin here, let's just take another look at the program at a glance. It's important

94
00:19:07.950 --> 00:19:27.870
Jim Donlon | NSF: to emphasize, and as you heard in the welcome that the program seeks to deliver on a national strategy to make sustained investments in Ai research in areas with the potential for long term payoff and national competitiveness and Ai. The emphasis on convergence foundational and use inspired. Research is clear. Each Institute has a clear role.

95
00:19:27.930 --> 00:19:34.730
Jim Donlon | NSF: I'm sorry has a critical goal, the significant advancement in the methods and understanding and foundational Ai.

96
00:19:34.740 --> 00:19:53.499
Jim Donlon | NSF: The concept of the ai institutes is that these Ai advances are sought in a use-inspired context, that is, with an emphasis on convergent research focused on important application areas, on scientific discovery and societal challenges which leads not only to foundational advances, but also ai-powered innovation for those centers.

97
00:19:53.790 --> 00:20:02.190
Jim Donlon | NSF: Ai institutes also leverage their scale and leadership position to nurture and grow the next generation of talent for our Ai power Future

98
00:20:02.960 --> 00:20:14.320
Jim Donlon | NSF: and Ai Institutes can be seen to lead in their chosen areas of Ai research and innovation, as they establish themselves in what we call nexus points for continuous growth in their collaborative efforts.

99
00:20:14.950 --> 00:20:23.730
Jim Donlon | NSF: We're now in our fourth issuance of the solicitations in this program, and this is for the next two rounds of awards to be made in two thousand and twenty, four and two thousand and twenty five

100
00:20:23.800 --> 00:20:33.260
Jim Donlon | NSF: in all. Previous rounds in the program we've been intentional in maintaining a consistent definition of what it is to be an Ai institute, and the solicitation continues that,

101
00:20:33.270 --> 00:20:43.990
Jim Donlon | NSF: however, while the identity of the program remains largely constant things do change over time, and you must be attentive to the solicitation, especially if you previously engaged the program

102
00:20:44.180 --> 00:21:01.789
Jim Donlon | NSF: in the revision notes for the solicitation. We call your attention to several notable changes. There will also be small revisions to solicitation text throughout, so as always, even if you're returning to the program, and have been familiar with previous solicitations. It's vital that you take the time to read the solicitation carefully

103
00:21:01.880 --> 00:21:05.790
this way. You can be most assured to be compliant and responsive you

104
00:21:05.800 --> 00:21:16.430
Jim Donlon | NSF: to highlight the most notable changes. First, as usual, we solicit proposals guided by several theme descriptions, and you've seen there are three, and we're going to expand on those in detail later in the Webinar

105
00:21:16.440 --> 00:21:23.949
Jim Donlon | NSF: second in each theme, and a step is joined by some funding partners, as you saw highlighted, and we'll get into more detail on that later.

106
00:21:23.960 --> 00:21:43.950
Jim Donlon | NSF: Third, we call your attention to some expansion we've added to the way we describe use-inspired research. This is to again attempt to amplify this idea. That use-inspired research is a fundamental mode or means by which we and we achieve our foundational ai and other research interests.

107
00:21:44.050 --> 00:21:58.679
Jim Donlon | NSF: And finally, as I mentioned up front, the solicitation covers the next two rounds, the next two years of anticipated awards. That means there's two different submission, and timelines and review timelines, and I'm going to cover that right now. Kind of at a glance one hundred and fifty

108
00:21:58.960 --> 00:22:16.069
Jim Donlon | NSF: mit ctl, and in the solicitation at the end of the program description. You're going to find a program, timeline reflected, divided into two groups, according to the year in which we're intending to make the awards. This timeline reflects the review, and so the election process as presented in that section of the solicitation, one

109
00:22:16.080 --> 00:22:24.290
Jim Donlon | NSF: in both groups, all themes. There is a requirement to submit preliminary proposals. October the Thirty first for astronomy, January the twelfth for the others.

110
00:22:24.300 --> 00:22:30.680
Jim Donlon | NSF: The intent is to return the results of preliminary proposal. Review for you within a month's time, as you can see here,

111
00:22:30.810 --> 00:22:49.590
Jim Donlon | NSF: full proposal due dates are February sixteenth, for astronomy, May seventeenth, for others Merit Review takes place, and either in either one of these timelines after proposal submission, Of course, during that period the program is likely to continue the practice of inviting some proposing teams for a reverse Site visit prior to final selection of institutes.

112
00:22:49.600 --> 00:23:01.590
Jim Donlon | NSF: Such teams are asked to choose key personnel to meet with the program officials via a video teleconference reflected. Here are the anticipated timelines for scheduling and conduct of these Rsvs. If there are changes to the schedule, two

113
00:23:01.600 --> 00:23:05.689
Jim Donlon | NSF: all the lead. Pis, who have submitted to the program will be notified by email

114
00:23:06.490 --> 00:23:15.939
Jim Donlon | NSF: awards made in astronomy are anticipated. October the first two thousand and twenty four, and the other themes have awards anticipated. To start June the first two thousand and twenty five.

115
00:23:17.550 --> 00:23:28.520
Jim Donlon | NSF: We now turn our attention to the activities that constitute an Ai Research Institute in this program. This is what we think of as the global scope of the program. The vision for the program is broad and ambitious.

116
00:23:28.530 --> 00:23:41.359
Jim Donlon | NSF: It is expected that each Ai Institute will pursue this vision in a unique way. But it's critical that you understand that every Ai Institute is selected for and evaluated on their impact in this full range of activities.

117
00:23:42.650 --> 00:23:56.580
Jim Donlon | NSF: The first function of institutes and critical to all proposals in this program is that Ai research Institutes must advance foundational Ai research. This means adding significant new knowledge and understanding to the Ai Research Central to your Institute's Vision

118
00:23:56.830 --> 00:24:18.039
Jim Donlon | NSF: Institutes might address new foundational Ai research priorities that arise from rapid advances in Ai, and the increasing ubiquity of Ai enabled technology or challenged disciplinary divisions that result from the current state of specialization, or promote the establishment of new science, engineering, and educational communities that better reflect the long term research needs for future Ai.

119
00:24:18.560 --> 00:24:29.739
Jim Donlon | NSF: It's very important to emphasize that the Institute proposals that do not describe a clear plan to achieve ambitious advances in foundational Ai research are not likely to be responsive to the program

120
00:24:30.480 --> 00:24:44.789
Jim Donlon | NSF: to understand the foundational Ai research that is in the scope of this program. We summarize here the definition of Ai that's provided in the solicitation. This definition helps to scope what the program is looking for. In seeking to quote significantly advanced research in Ai.

121
00:24:45.680 --> 00:25:04.239
Jim Donlon | NSF: In a broad sense, Ai is concerned with understanding the mechanisms underlying thought and intelligent behavior and their implementation of machines. What we often refer to as Core Ai research at Nsf. Are methods pertaining to the learning, abstraction, and inference considered essential for the implementation of intelligent behavior and machines.

122
00:25:04.370 --> 00:25:15.719
Jim Donlon | NSF: Corey Ah, Research also includes the study of architectures intended to directly manifest that behavior, including general architectures for intelligence, integrated intelligent agents, and multi-agent systems.

123
00:25:17.150 --> 00:25:24.220
Jim Donlon | NSF: The modeling and implementation of intelligent behavior and machines, began as and has continued to be a multidisciplinary endeavor.

124
00:25:24.430 --> 00:25:33.869
Jim Donlon | NSF: Many computational models of intelligence have drawn inspiration from living systems, drawing from biologically inspired computational neuroscience and behavioral and cognitive science

125
00:25:34.220 --> 00:25:44.130
Jim Donlon | NSF: computer vision and human language. Technologies have a high overlap with artificial intelligence, and as much as they provide methods for the perceptual and communications, capabilities critical to intelligent systems.

126
00:25:45.020 --> 00:25:53.400
Jim Donlon | NSF: Robotics is closely related to, but not identical to Ai. Robotics can provide the environment critical for intelligence systems to be able to act upon the world

127
00:25:53.410 --> 00:26:09.330
Jim Donlon | NSF: while an embodied Ai may be a robot. The solicitation focuses on the Ai aspect of that research. For example, the solicitation does not include in its scope development that is mainly focused on tele-operated robots or robots that repeat programmed instructions in controlled environments

128
00:26:11.330 --> 00:26:19.519
Jim Donlon | NSF: moving on in this list of six to zero. The second is that Ai research institutes will continue, or will conduct use-inspired research

129
00:26:19.530 --> 00:26:32.110
Jim Donlon | NSF: in the Ai Institute's program this means that a use-inspired context. Both informs the foundational ai research that we just discussed and drives innovations in related sectors through focus on use cases or applications.

130
00:26:32.490 --> 00:26:41.689
Jim Donlon | NSF: The solicitation emphasizes that this should be conveyed in terms of clear and compelling bottles can enhance knowledge, transfer, and create the potential for shared community infrastructure.

131
00:26:41.700 --> 00:26:57.989
Jim Donlon | NSF: It is important to understand that Use-inspired research is a means of achieving the first goal of foundational Ai, further expanding the context and value of the Institute's vision and the foundational Ai goals are a means for achieving transformative advances in the use inspired sector.

132
00:26:58.000 --> 00:27:06.510
Jim Donlon | NSF: A compelling use inspired context should balance the application of Ai with the advancement of foundational Ai in a way that is appropriate to your institute's vision,

133
00:27:08.130 --> 00:27:26.939
Jim Donlon | NSF: use-inspired research refers to basic research that has use for society in mind. The conceptual depiction on the right shows the broad potential for a range of use Inspired invertibles, if you will, to inform and benefit from a wide range of potential foundational ai priorities.

134
00:27:27.130 --> 00:27:40.870
Jim Donlon | NSF: Institutes established in the program to date have found many unique intersections in this conceptual space. As before. We've taken inspiration from partners and the larger community to invite you. The proposing community to come at this from different directions

135
00:27:40.880 --> 00:27:49.280
Jim Donlon | NSF: in the astronomy and materials themes. In the solicitation we'll soon see that their intended domains of scientific discovery are largely defined in the scope of the theme.

136
00:27:49.370 --> 00:27:54.439
Jim Donlon | NSF: Yet consider, there might be room for more institute specific visioning within those pursuits

137
00:27:54.580 --> 00:28:02.919
Jim Donlon | NSF: in strengthening Ai, where the theme scope is strict only in terms of Ai goals. You'll see that the potential for use inspire and inspiration and impact is limitless.

138
00:28:06.480 --> 00:28:13.740
Jim Donlon | NSF: The goal in this program is that each institute conducts foundational research within uses by our context in a deeply integrated way.

139
00:28:13.790 --> 00:28:19.690
Jim Donlon | NSF: So ideally the results from use. Inspired application areas are generalized and even made foundational.

140
00:28:19.700 --> 00:28:36.959
Jim Donlon | NSF: Um! But before we leave this ah, this notion of foundational Ai, and you suspended, I want to address a question we're sometimes asked about how to select an appropriate balance in an institute set of goals between foundational Ai and other domains.

141
00:28:36.970 --> 00:28:45.750
Jim Donlon | NSF: We want to warn against inappropriately putting these concepts in opposition to one another, and instead recall the virtuous cycle we aspire to.

142
00:28:45.760 --> 00:28:50.789
Jim Donlon | NSF: Not likely competitive, for the program is a proposal in which Ai is merely applied to the domain,

143
00:28:50.800 --> 00:28:55.780
Jim Donlon | NSF: or even in which Ai goals are foundationally yet seem not well integrated with the domain.

144
00:28:56.080 --> 00:29:13.020
Jim Donlon | NSF: In the introduction to the solicitation we identified the preferred convergent approach as being at the very heart of the program's identity. Use-inspired research is, quote an expected modality of or means by which we wish to see you propose the foundational Ai advances one.

145
00:29:14.290 --> 00:29:17.700
Jim Donlon | NSF: The program description expands on the program. Requirement

146
00:29:17.950 --> 00:29:21.380
Jim Donlon | NSF: defining the virtuous cycle in a non-superficial way

147
00:29:21.420 --> 00:29:31.959
Jim Donlon | NSF: the use-inspired context reveals the opportunities for foundational Ai advances, and those quote foundational Ai advances in term contribute to the domain.

148
00:29:32.740 --> 00:29:38.180
Jim Donlon | NSF: Finally, keep in mind that Nsf. And reviewers will evaluate proposals and later institutes.

149
00:29:38.200 --> 00:29:46.200
Jim Donlon | NSF: How well the activity leverages, this use inspired context consistent with your program, with your Institute vision in the program vision

150
00:29:48.890 --> 00:29:55.169
Jim Donlon | NSF: moving on further with these sixties that are out, and the third is to grow the next generation of talent.

151
00:29:55.250 --> 00:30:06.649
Jim Donlon | NSF: Ai Researcher. A research institutes are expected to maximize their unique position to grow the next generation of talent for a diverse, well-trained workforce that will provide new discoveries and leadership.

152
00:30:07.100 --> 00:30:29.449
Jim Donlon | NSF: Specifically, these institutes should leverage the visionary nature of their research book guide to drive new and innovative education and development tailored toward the nation's, undergraduates, graduate students and post-doctoral researchers as well as through community colleges and skilled technical workforce, training, and other opportunities that advance knowledge and education of Ai, including public understanding of Ai.

153
00:30:29.970 --> 00:30:42.569
Jim Donlon | NSF: This could include an innovative pedagogy in structural materials, advanced learning technologies, project-driven training, cross-disciplinary and collaborative research industry partnerships new career pathways,

154
00:30:42.630 --> 00:30:59.770
Jim Donlon | NSF: institutes should offer broad, deep and diverse experiences to build the next generation of the Ai workforce, with a focus on broadening participation among groups that are starkly underrepresented in science and engineering, including women and people with disabilities and people from other minority minorities. Groups,

155
00:31:00.560 --> 00:31:12.770
Jim Donlon | NSF: institutes must propose meaningful plans for these activities, and the solicitation describes required sections on education and workforce development, and the broadly participation plan that focus on education and workforce development.

156
00:31:12.790 --> 00:31:25.270
Jim Donlon | NSF: The caution proposals, however, are again seeing these as separate from the other proposed activities. Rather education, training, and inclusion plans should be well integrated with the other research and coordination activities described in the proposal

157
00:31:25.460 --> 00:31:28.640
Jim Donlon | NSF: forming a hole that is greater than the sum of its parts.

158
00:31:29.950 --> 00:31:35.549
Jim Donlon | NSF: Fourth desitteratom. We shorthand here as institutes should be multi-disciplinary

159
00:31:35.560 --> 00:31:47.290
Jim Donlon | NSF: ai research institutes are coherent, multidisciplinary groups of scientists engineers and educators appropriate for a large-scale long-term research agenda for the advancement of Ai and the field ended. Ai.

160
00:31:47.940 --> 00:31:56.530
Jim Donlon | NSF: The multi-disciplinary nature of these institutes will catalyze foresight and adaptability beyond what is possible in the single research projects

161
00:31:56.900 --> 00:32:13.430
Jim Donlon | NSF: this objective echoes of program's observation that Ai itself has traditionally been highly multidisciplinary, and that use inspired research requires synergy among a group of researchers enabling transformative advances in Ai related sectors and the interfaces between those areas.

162
00:32:13.880 --> 00:32:21.769
Jim Donlon | NSF: In reviewing institute proposals, reviewers and panels look for signs that the individual projects of an institute will be meaningfully integrated,

163
00:32:22.000 --> 00:32:31.450
Jim Donlon | NSF: resulting in contributions beyond the some of the individual projects. Significant integration of multidisciplinary research can be a significant basis of confidence in this regard.

164
00:32:33.790 --> 00:32:40.980
Jim Donlon | NSF: Ai. Research institutes are expected to be comprised of multiple organizations working together to create new research capabilities.

165
00:32:40.990 --> 00:32:53.740
Jim Donlon | NSF: Nsf. And partner organizations seek to grow the network of National Ai Research institutes and lead organizations distributed throughout the country to grow new centers of Ai leadership. Leveraging existing centers of excellence as appropriate

166
00:32:54.020 --> 00:33:01.190
Jim Donlon | NSF: institutes are strongly encouraged to include organizations that can directly contribute to Nsf's commitment to broadening participation

167
00:33:01.200 --> 00:33:14.050
Jim Donlon | NSF: by engaging a diverse globally engaged research community integrating research with education and building capacity and expanding efforts to broaden participation from underrepresented groups in diverse institutions across all geographic regions.

168
00:33:14.570 --> 00:33:25.399
Jim Donlon | NSF: We going to draw your attention to the importance of demonstrating that your proposal be greater than the sum of its parts here, by meaningfully leveraging and integrating diverse organizations.

169
00:33:26.080 --> 00:33:33.939
Jim Donlon | NSF: In addition to a lead pi suitable to lead an institute, and each institute must also be staffed with a managing director or a project manager.

170
00:33:34.220 --> 00:33:50.630
Jim Donlon | NSF: And finally, it's important to emphasize, while an external advisory board will be required for an institute. Please do not contact, invite, or identifying your proposal, any potential eab members. The step should be taken after the merit, review, and selection process for the sake of coi management.

171
00:33:52.100 --> 00:33:58.240
Jim Donlon | NSF: Sixth and final one is that all Ai institutes are expected to be nexus points for collaborative efforts.

172
00:33:58.250 --> 00:34:10.069
Jim Donlon | NSF: By nexus point we refer to the role that an institute can play in creating an organization that encourages the continuing growth of collaborations, especially with external partners that can further enhance the impact of the Institute.

173
00:34:10.409 --> 00:34:21.880
Jim Donlon | NSF: The nexus point function of the program is not a mere state of being. In other words. When you have assembled your multi-organizational Ai institute you, don't point to it and say, look at our nexus point, one

174
00:34:21.889 --> 00:34:28.820
Jim Donlon | NSF: being an access point is an active set of priorities, programs, mechanisms that are aware of an Ai research institute.

175
00:34:28.830 --> 00:34:48.220
Jim Donlon | NSF: Several things establishes itself as a thought leader and research leader in their chosen area and pursues the continuing growth of collaborations with external partners to bring together people ideas, problems, and technical approaches for maximum and and and extended impact beyond the members and the boundaries of the Institute itself.

176
00:34:48.929 --> 00:35:00.500
Jim Donlon | NSF: Consider even a proposal. Time, what kinds of priorities, programs, and mechanisms your Institute could use to connect with new partners and create organizational collaborations and linkages with a range of organizations.

177
00:35:03.190 --> 00:35:14.310
Jim Donlon | NSF: Having emphasized the essential nature of all Ai institutes, we are now ready to turn our attention to the three specific themes or high priority areas in which proposals are solicited. For the next two rounds of awards

178
00:35:14.320 --> 00:35:26.799
Jim Donlon | NSF: there are themes representing a subset of research areas that an Asf and partner support, and which were determined to be both an immediate priority for funding partners, and for which the resource community was judged to be ready to be responsive this year,

179
00:35:26.830 --> 00:35:31.470
Jim Donlon | NSF: and a seps estimating approximately five rewards, as shown here.

180
00:35:32.230 --> 00:35:46.390
Jim Donlon | NSF: And so we'll let the theme leads. Say more about this. Ah, but before I invite the theme leads to expand on these, let me just address a question that comes up about submitting a proposal that might respond to multiple themes, one hundred and fifty.

181
00:35:46.400 --> 00:35:50.910
Jim Donlon | NSF: Our answer here is, while it's possible, of course, to scope your proposal this way.

182
00:35:50.920 --> 00:36:03.129
Jim Donlon | NSF: Your proposal must be submitted to one theme as the primary one you're submitting to. So such overlaps may not hurt your responsiveness. If you've done nothing to compromise or focus on that primary theme you submitted to.

183
00:36:03.610 --> 00:36:18.709
Jim Donlon | NSF: However, consider each theme to be an independent funding pool. As you think about the relevance to multiple thieves. Viewed that way. You can see that it's not necessarily the case. That special effort to make the proposal relate to the goals of another theme will make it more likely to succeed overall.

184
00:36:18.720 --> 00:36:36.030
Jim Donlon | NSF: It's most effective to respond as fully as possible to the scope of the theme you address. Relevance to other themes might be identified in the proposal text for Keywords section of the project. Summary, if you like, and where such level relevance is particularly strong and assessment partners may co-review such proposals in those secondary themes

185
00:36:36.710 --> 00:36:53.689
Jim Donlon | NSF: as always. It's a good idea to consult with the respective program contacts for each theme to get advice about your specific situation. So with that bit of preamble out of the way i'd like to present. Ah, turn the presentation over to my colleagues. I want to start with Andreas Berlin from the division of astronomy who is going to describe the one

186
00:36:53.700 --> 00:36:54.759
Jim Donlon | NSF: press.

187
00:36:55.070 --> 00:36:56.489
Andreas Berlind (NSF/AST): Thank you, Jim.

188
00:36:56.500 --> 00:37:13.020
Andreas Berlind (NSF/AST): Hi, Everyone uh I'm Andreas Berlin. I'm. A program director in the Astronomy Division here at the Nsf. Ah, and I'm. Leading the astronomy theme of the solicitation. This is the first time that that our division has participated in this program, and and we're very excited to be part of it.

189
00:37:13.300 --> 00:37:21.009
Andreas Berlind (NSF/AST): So as as a Dr. Punch said earlier, astronomy is a research field that is extraordinarily data rich.

190
00:37:21.020 --> 00:37:34.039
Andreas Berlind (NSF/AST): Our data is coming in rapidly from lots of telescopes around the world, and future experiments will accelerate this dramatically. The data cover a wide range of wavelengths of light will come at high temporal cadence.

191
00:37:34.360 --> 00:37:40.799
Andreas Berlind (NSF/AST): Addition, theoretical models that are attempting to simulate the data are themselves producing massive volumes of data.

192
00:37:41.840 --> 00:37:51.179
Andreas Berlind (NSF/AST): Ai technology is rapidly entering our research field, and more and more will be essential for mining, analyzing, and extracting knowledge from these data.

193
00:37:51.360 --> 00:37:58.259
Andreas Berlind (NSF/AST): But solving the biggest challenges will require advances in foundational ai that are inspired by them. The

194
00:37:58.430 --> 00:38:05.480
Andreas Berlind (NSF/AST): this is a challenge in our field, because there isn't enough interdisciplinary collaboration between astronomers and Ai researchers

195
00:38:05.490 --> 00:38:08.629
Andreas Berlind (NSF/AST): in this, in the realm of education,

196
00:38:08.640 --> 00:38:14.899
Andreas Berlind (NSF/AST): Ai. Education within our astronomy community is not done uniformly, and is not always informed, vested in the art.

197
00:38:15.150 --> 00:38:23.270
Andreas Berlind (NSF/AST): So to address these challenges. The Nsf. In partnership with the Simon's foundation. Aim to fund up to two Institutes,

198
00:38:23.280 --> 00:38:35.830
Andreas Berlind (NSF/AST): and our hope is that these institutes will tackle important and research challenges in astronomy through multidisciplinary disciplinary collaboration, and will serve as hubs of knowledge and education and ai for the broader community

199
00:38:36.320 --> 00:38:37.950
Andreas Berlind (NSF/AST): next slide, please.

200
00:38:39.550 --> 00:38:58.689
Andreas Berlind (NSF/AST): So. Ah, i'd like to just go over. Ah sort of our themes version of the desirado that the gym I went over earlier. Ah! Competitive proposals to this theme will identify important challenges in astronomy research that that will be addressed by the proposed institute. One,

201
00:38:58.700 --> 00:39:07.759
Andreas Berlind (NSF/AST): as I'll say again, later, the solicitation is broad, so many areas are possible.

202
00:39:07.770 --> 00:39:12.450
Andreas Berlind (NSF/AST): But the research plan should be appropriately ambitious. Given the skill of these grants,

203
00:39:13.260 --> 00:39:26.869
Andreas Berlind (NSF/AST): Equally important is the proposals should should describe the advances in the Foundational Ai that will be needed to tackle these challenges, and the Foundational Ai and astronomy. Research should be well integrated with each other

204
00:39:27.510 --> 00:39:37.279
Andreas Berlind (NSF/AST): in this vein. Proposals that describe application of off-the-shelf Ai models to astronomy problems are not likely to be responsive to the solicitation

205
00:39:38.610 --> 00:39:57.170
Andreas Berlind (NSF/AST): competitive proposals will also present a plan for the proposed Institute to serve as a hub of Ai knowledge and education for the broader astronomical community. We would really like to see these two institutes that emerge from this competition have a broad impact and elevate the level of Ai research with an astronomy,

206
00:39:57.290 --> 00:40:07.509
Andreas Berlind (NSF/AST): and we encourage proposals to include activity that reaches researchers and institutions that do not have easy access to cutting edge data on methods.

207
00:40:08.270 --> 00:40:18.289
Andreas Berlind (NSF/AST): And finally, proposals should clearly describe the organization and management structure of the Institute, including the roles of different organizations and team members,

208
00:40:18.300 --> 00:40:28.710
Andreas Berlind (NSF/AST): and the proposed collaborations that come forth should be motivated by the science and Education goals of the Institute, and should add up to more than the sum of their parts, as as Jim mentioned before.

209
00:40:31.870 --> 00:40:33.429
Andreas Berlind (NSF/AST): So um

210
00:40:33.910 --> 00:40:39.720
Andreas Berlind (NSF/AST): i'd like to just go over Briefly, a few commonly asked questions that we've encountered so far,

211
00:40:39.740 --> 00:40:44.710
Andreas Berlind (NSF/AST): So the most common question we've gotten is, what areas of astronomy are eligible?

212
00:40:45.320 --> 00:41:04.989
Andreas Berlind (NSF/AST): And The answer is, it's very broad. All areas of astronomy that that we support in our division are are eligible, and and that really covers the whole range from planetary solar, stellar, Galactic Extra Galactic Cosmology, interdisciplinary areas like multi messenger astronomy and theory.

213
00:41:05.000 --> 00:41:14.550
Andreas Berlind (NSF/AST): So this is the station is on purpose intended to be broad and we want to see the best ideas from all parts of the field.

214
00:41:15.400 --> 00:41:19.199
Andreas Berlind (NSF/AST): What about research that touches on neighboring fields like physics?

215
00:41:19.270 --> 00:41:27.639
Andreas Berlind (NSF/AST): Um, research, activity and other fields. Is Okay, as long as the main focus is on on uh on astronomy.

216
00:41:28.500 --> 00:41:31.069
Andreas Berlind (NSF/AST): What about space-based astronomy

217
00:41:31.080 --> 00:41:32.390
Andreas Berlind (NSF/AST): Because this is something that

218
00:41:32.400 --> 00:41:39.890
Andreas Berlind (NSF/AST): the Nsf. And past and other solicitations has not supported as much as ground basis

219
00:41:39.900 --> 00:41:42.550
Andreas Berlind (NSF/AST): in this elicitation. This is not restricted

220
00:41:42.740 --> 00:41:48.199
Andreas Berlind (NSF/AST): The proposed ai work can involve space-based as well as ground-based research and data.

221
00:41:48.440 --> 00:41:54.049
Andreas Berlind (NSF/AST): The main thing is that the proposed research is sufficiently ambitious. Given the size and scale of the awards,

222
00:41:55.550 --> 00:42:01.169
Andreas Berlind (NSF/AST): Is it better for a proposal to tackle one astronomy challenge in depth or touch on a broad set of topics.

223
00:42:01.440 --> 00:42:11.909
Andreas Berlind (NSF/AST): The solicitation does not restrict this. What's important is that the research is sufficiently impactful, whatever it is, and is well integrated with foundational Ai work. The

224
00:42:12.210 --> 00:42:19.219
Andreas Berlind (NSF/AST): and finally do. The Nsf. And the Simon's foundation have different priorities that need to be balanced somehow.

225
00:42:19.230 --> 00:42:26.999
Andreas Berlind (NSF/AST): The answer is, no. The theme, again, is intended to be broad. Both funding partners welcome all ideas from the research community.

226
00:42:30.940 --> 00:42:37.650
Andreas Berlind (NSF/AST): So there are a couple of specific things that relate to the partnership with Simons that we should. We should address

227
00:42:37.700 --> 00:42:54.169
Andreas Berlind (NSF/AST): mit ctl and um. So in terms of the budget. Half of the proposed budget, which is up to ten million dollars, needs to be prepared for the Simon's foundation, following instructions from the science foundation, and one

228
00:42:54.180 --> 00:42:59.600
Andreas Berlind (NSF/AST): that part of the budget will be submitted as a supplementary document in the Nsf. Proposal.

229
00:42:59.750 --> 00:43:10.159
Andreas Berlind (NSF/AST): So the budget will be split into with activities clearly delineated as on the on the left side and on the Simon side, in order to prepare the budget.

230
00:43:10.640 --> 00:43:22.429
Andreas Berlind (NSF/AST): Um keep. Keep in mind that the Simon's foundation has a different indirect rate policy than the Nsf. And you need to look at that to adhere to it.

231
00:43:23.290 --> 00:43:29.089
Andreas Berlind (NSF/AST): Um. Another uh Another result of the partnership with Simon is that,

232
00:43:29.370 --> 00:43:38.440
Andreas Berlind (NSF/AST): in addition to the Nsfpi meetings that will happen, the Silence Foundation will organize and found annual Api meetings at the Simon's foundation in New York City.

233
00:43:38.890 --> 00:43:42.900
Andreas Berlind (NSF/AST): Um. And we'll have reporting requirements.

234
00:43:43.660 --> 00:43:47.709
Andreas Berlind (NSF/AST): The review of proposals will be conducted by the Nsf.

235
00:43:48.330 --> 00:43:57.539
Andreas Berlind (NSF/AST): But Simon's foundation personnel may observe panels and participate in the review process, and we expect to make a word decisions together.

236
00:43:58.280 --> 00:44:06.040
Andreas Berlind (NSF/AST): Uh one thing I really want to emphasize is that, uh, despite the fact that the budget will will have these two components;

237
00:44:06.190 --> 00:44:14.529
Andreas Berlind (NSF/AST): that both agencies, both the Msf. And the Simons foundation will evaluate the proposal as a whole, and the budget as a whole.

238
00:44:14.550 --> 00:44:24.290
Andreas Berlind (NSF/AST): The way in which the the budget is split will have bookkeeping and project management implications, but is not intended to affect proposal. Review.

239
00:44:25.120 --> 00:44:26.390
Andreas Berlind (NSF/AST): It's A.

240
00:44:28.270 --> 00:44:31.359
Andreas Berlind (NSF/AST): And finally,

241
00:44:31.530 --> 00:44:42.190
Andreas Berlind (NSF/AST): I want to address the issue of including people who are affiliated with Simon's foundation on the proposal. For example, people who are affiliated with a flattering institute.

242
00:44:42.260 --> 00:45:00.309
Andreas Berlind (NSF/AST): And um first of all, Ah! The participation of these people has some restrictions. First of all, you should check with the silence, foundation, and ensure compliance with their policy on eligibility of Simon's employees for external funding one hundred and fifty.

243
00:45:00.330 --> 00:45:19.960
Andreas Berlind (NSF/AST): But as the solicitation solicitation, states, people affiliate with Simons may not participate in a proposal to this theme in their capacity, as Ah Simon's employees however, if they have an appointment at a non-signments, and indeed, for example, a university.

244
00:45:20.530 --> 00:45:24.459
Andreas Berlind (NSF/AST): They may participate in a proposal in that capacity.

245
00:45:24.510 --> 00:45:25.779
Andreas Berlind (NSF/AST): It's

246
00:45:26.280 --> 00:45:29.879
Andreas Berlind (NSF/AST): in an unfunded basis, meaning they cannot draw a salary.

247
00:45:29.930 --> 00:45:31.240
Andreas Berlind (NSF/AST): Ah,

248
00:45:31.540 --> 00:45:38.989
Andreas Berlind (NSF/AST): you have questions about this, or you know, some specific case, or any of the other

249
00:45:39.000 --> 00:45:50.079
Andreas Berlind (NSF/AST): things I mentioned, I feel free to put a question in the in the Q. And A. I'm. Also happy to meet with you separately to discuss your proposal. Thank you.

250
00:45:55.630 --> 00:46:03.040
Jim Donlon | NSF: Yeah. With that we're going to hand straight over to Senator, who's going to speak for Ai for discovering materials. Research.

251
00:46:03.050 --> 00:46:04.889
Serdar Ogut: Thanks very much, Jim.

252
00:46:04.910 --> 00:46:20.529
Serdar Ogut: Hello, everyone. I'm uh Sara. I'm a program director in the condensed matter and material security program of the division of materials research at Nsf. And one of the contacts along with Sylvia Spangler from size for theme. Two of the soil station,

253
00:46:20.560 --> 00:46:34.920
Serdar Ogut: Um, as Jim already mentioned earlier. This is the fourth round of it. Ni. A research is useful stations, and the first one like astronomy, in which the vision of materials, research will directly participate.

254
00:46:34.930 --> 00:46:56.980
Serdar Ogut: Dmr's participation in this solicitation is based on the recognition that Ai has been playing an increasingly important role in materials research as more and more material scientists, physicists, chemists are realizing the power of Ai technologies to revolutionize materials, discovery and understanding.

255
00:46:56.990 --> 00:47:14.060
Serdar Ogut: This team builds on the umr's, current impactful efforts on the material genome initiative, Mgi and expands the scale and integration of scientific challenges to enable ai-based transformative advances in materials research.

256
00:47:14.130 --> 00:47:35.410
Serdar Ogut: Ah, We We expect that a successful um material of Ai Institute will make breakthrough advances in materials. The story and understanding enable new ai-based capabilities, push the boundaries of foundational Ai. While being responsive to societal challenges and industrial needs,

257
00:47:35.750 --> 00:47:45.809
Serdar Ogut: as mentioned earlier, Intel will be providing partial support for this Institute theme, and i'll be talking about that briefly, in a little while,

258
00:47:45.820 --> 00:47:59.060
Serdar Ogut: overall proposals that advance both foundational Ai and fundamental materials. Research in domains that are supported by Dmr will be most responsive to this theme

259
00:47:59.070 --> 00:48:27.700
Serdar Ogut: I emphasize fundamental materials research to distinguish it from areas that are typically within the engineering domains. In in a sense, if you would, uh for the materials portion of this, if you would normally uh submit this proposal to programs. Uh in the engineering directorate. Um. That may not be. Uh, they may not be the best for this. Uh the next slide, please.

260
00:48:31.790 --> 00:48:57.539
Serdar Ogut: Um. As I mentioned earlier, we envisioned the materials. Ai Institute to expand Themr's ongoing efforts in Mgi. To those who may not be familiar material Genome initiative is a Federal multi-agency, initiative launched in two thousand and eleven for discovering manufacturing and deploying advanced materials twice as fast and at a fraction of the cost compared to a traditional method

261
00:48:57.550 --> 00:49:18.679
Serdar Ogut: uh the Dmrf program, short for uh designing materials to revolutionize and engineer. Our future has been running for the last ten years, as Dmr's primary response to Mgi. I should also mention here, in passing, that Bmr. Is a major partner in the Mrf. But there are other directories and programs at Nasa that participate in it as well

262
00:49:19.280 --> 00:49:38.779
Serdar Ogut: in the two thousand and twenty one mgi strategic plan, harnessing the power of materials. Data has been identified as one of the primary goals. The objectives and the skull are centered on ai-driven materials R. And D. For retur of discovery and manufacturers.

263
00:49:38.790 --> 00:49:50.240
Serdar Ogut: The new materials Ai. Institute that will be funded in this, so station can be viewed as part of the Emr's continued efforts in advancing the Mgi approach and Bishop

264
00:49:50.520 --> 00:49:52.029
Serdar Ogut: next slide, please.

265
00:49:53.770 --> 00:50:06.119
Serdar Ogut: So if you look at the description of the material steam in the Solar station. You'll realize that it is material and application, domain agnostic, and this is by design,

266
00:50:06.130 --> 00:50:34.520
Serdar Ogut: by specifically not focusing on some subset of material problems or resource models or tools. We wanted to solicit ideas from the entire materials research community on what burning materials, research problems, they think would benefit the most from recent advances in foundational Ai, and in turn push the boundaries of Foundational Ai to address challenges and opportunities in materials. Research,

267
00:50:34.530 --> 00:51:04.199
Serdar Ogut: however, we wanted to provide some guidance here actually more towards stimulating thought from research at the intersection of Ai and materials. Hence we provided the list that you so. Ah! That you probably saw in the association of five potential lines of research, and these are globally multi-moral. Ah, data integration and data set development foundational Ai denses driven by materials research

268
00:51:04.510 --> 00:51:13.390
Serdar Ogut: first synthesis to synthesis at scale human augmented materials, design and interpretable materials. Ai:

269
00:51:13.740 --> 00:51:36.410
Serdar Ogut: The proposals that will be responsive to this theme can focus on any area or areas of fundamental materials. Research aligned with the division of materials, research and employ various research models from a practical perspective, even though a particular institute may focus on some subset of material problems or modes of research.

270
00:51:36.420 --> 00:51:49.519
Serdar Ogut: The potential generalizability of this focus choice to other materials, problems and research modes is likely to align it with the Ai ambition to generalize from data

271
00:51:50.360 --> 00:51:51.979
Serdar Ogut: next slide, please.

272
00:51:54.130 --> 00:52:11.460
Serdar Ogut: So here I want to thank again on the industry funding partner intel for joining Nsf. In this important Ai research. Ah, priority, this slide summarizes the the nature of Intel's participation. You can find more details in the source station.

273
00:52:11.470 --> 00:52:28.759
Serdar Ogut: Um. The proposals that are respond to this theme should be careful not to add conflicts of interest with this partner in the ways shown here in particular. Ah! I know that Intel will not participate in world zero panels, or may not review.

274
00:52:28.770 --> 00:52:40.670
Individuals affiliated with Intel are not allowed to participate in proposals to them to in their capacity at Intel, but they may be able to do so in their capacity, and and another organization

275
00:52:40.680 --> 00:52:51.530
Serdar Ogut: proposers should not contact intel about proposals to the scene, or make any prior arrangements for collaborations in the scheme during the proposal,

276
00:52:51.760 --> 00:53:06.160
Serdar Ogut: after the award is processed in town may offer collaboration opportunities to the Funding Institute in the theme, including software datasets, other computing of infrastructure and or researchers in residents.

277
00:53:06.650 --> 00:53:36.640
Serdar Ogut: Ah, it is um ah discussing one of the common questions that we have received so far that pertains to whether the involvement of Intel Ah! As an industry ah funding partner would mean that the proposed research should be um, or have a projection on semiconductor research. As I mentioned earlier, this is not the case at all. Partnership of Intel does not restrict the materials. Research themes of this Ah Institute proposal. As I mentioned

278
00:53:36.650 --> 00:53:52.200
Serdar Ogut: earlier, We're interested in soliciting ideas from the entire materials, community for research that is aimed at advancing materials, discovery, and understanding while advancing the frontiers of foundation of they are. And with that that's the two hectares.

279
00:53:54.050 --> 00:54:01.059
Hector Munoz-Avila | NSF: So good afternoon, everyone. My name is Sector Minosavira. I am a program director in size, and the team lead fourteen three.

280
00:54:01.130 --> 00:54:15.769
Hector Munoz-Avila | NSF: Let's go over the motivation for this theme. Ai assistance. Build with the layer architecture strain in massive, massive datasets, have become increasingly capable of producing useful and impressing outlets

281
00:54:26.360 --> 00:54:28.950
Hector Munoz-Avila | NSF: for them. More decisions can be brutal.

282
00:54:39.470 --> 00:54:54.230
Hector Munoz-Avila | NSF: Mit Ctl and the Isis of the future will need to be a strengthened if they are to be as robust as we would like. And if we are to keep such technology well aligned with societal intended uses nexus

283
00:54:58.760 --> 00:55:02.300
Hector Munoz-Avila | NSF: basic requirements for team. Three

284
00:55:02.320 --> 00:55:16.099
Hector Munoz-Avila | NSF: Institute funded under this theme, mostly not lead advances in theory methods, or iterative approaches that are stranded. Ai. In all three of the goals listed below, Grounding

285
00:55:16.230 --> 00:55:17.790
Hector Munoz-Avila | NSF: instructibility

286
00:55:17.800 --> 00:55:18.899
Hector Munoz-Avila | NSF: in alignment.

287
00:55:26.230 --> 00:55:32.919
Hector Munoz-Avila | NSF: Grounding means that sisters must understand the concepts they reason over and operate with

288
00:55:33.200 --> 00:55:41.079
Hector Munoz-Avila | NSF: rounding allows Ai systems to demonstrate connection between yourself and the after concert, that they operate with

289
00:55:53.550 --> 00:56:07.060
Hector Munoz-Avila | NSF: instructibility enable systems that can be proven experimentally to change their behavior appropriately in response to explicit feedback provided by even non expert users.

290
00:56:14.330 --> 00:56:15.799
Hector Munoz-Avila | NSF: Next a slide, please.

291
00:56:19.320 --> 00:56:34.699
Hector Munoz-Avila | NSF: By alignment we mean that sisters must be judged by how well their operations align with expectations of objective truth in a domain, and correspond to societal expectations and human intentions in their operations.

292
00:56:35.300 --> 00:56:36.879
Hector Munoz-Avila | NSF: Next slide, please

293
00:57:08.010 --> 00:57:09.429
Hector Munoz-Avila | NSF: next slide, please.

294
00:57:23.360 --> 00:57:38.329
Hector Munoz-Avila | NSF: This is like summarize the nature of the partners of participation. More details are in the solicitation in a nutshell Proposal responding to this team should be careful not to have conflict of interest with these partners in the way shown here. One

295
00:57:38.340 --> 00:57:44.980
Hector Munoz-Avila | NSF: seriously situation for full details and contact us with any questions You may have next a slide, please.

296
00:57:50.270 --> 00:58:02.899
Hector Munoz-Avila | NSF: As we mentioned earlier, the office of the Undersecretary of Defense of Research and engineering at least have partnered with Nsf to potentially contribute funding to institutions funded under this award one.

297
00:58:11.120 --> 00:58:15.580
Hector Munoz-Avila | NSF: They are to where you can communicate with Nsf. And partners about this.

298
00:58:15.750 --> 00:58:34.720
Hector Munoz-Avila | NSF: If you wish to identify aspects of the your team. Free proposal that you believe are relevant to one or more partners job. The option to submit a supplementary document, both preliminary and proposal and stages to indicate relevance to your of your proposal to one or more.

299
00:58:34.730 --> 00:58:37.969
Hector Munoz-Avila | NSF: See solicitation for full instruction details.

300
00:58:38.050 --> 00:58:48.889
Hector Munoz-Avila | NSF: If you wish to indicate an objection to receiving funding in a team three proposal that are the units. From one of these two partners you may submit an optional single copy document.

301
00:59:01.800 --> 00:59:17.169
Hector Munoz-Avila | NSF: Finally, We know that here that either type of document, the strive here submitted with a proposal either preliminary or full, pertains only to that proposal. Documents submitted with the preliminary proposal, do not pertain to the full proposal

302
00:59:17.180 --> 00:59:26.329
Hector Munoz-Avila | NSF: that you may change or clarify your position to these issues at the full proposal stage, and with that head back to Gene. Thank you

303
00:59:26.950 --> 00:59:39.299
Jim Donlon | NSF: mit Ctl: and thank you, Hector. Thank you, colleagues. So now that we've finished looking at the individual themes, we're almost ready for Q. And A. We've been typing some answers in the module, so please continue to submit your questions there two hundred and fifty

304
00:59:39.310 --> 01:00:08.710
Jim Donlon | NSF: Um. Where the themes are concerned. Our program context will advise you on questions and responsiveness to the theme. In compliance with the solicitation. If you make contact with them, we can't give you any judgments or advice about the competitiveness of your idea. However, we encourage you to make contact with us prior to submitting your preliminary proposal, especially if you have questions about the response of this and your idea, we prefer that you email your theme-specific questions to all program contacts listed for that theme, so that we can give you a coordinated and complete response

305
01:00:08.830 --> 01:00:14.739
Jim Donlon | NSF: for feedback on the responsiveness of a proposal idea. You should include a project summary of up to two pages

306
01:00:14.870 --> 01:00:42.299
Jim Donlon | NSF: after the preliminary proposal. Deadline. It's more awkward to give you a really targeted advice because we've sent you. We've sent you feedback on the response responsiveness of your preliminary. So those program specific inquiries should be limited to request for clarification of that feedback that you received on the prelim these requests for clarification must be submitted to the theme contacts. And again, please address all the theme. Contacts that are listed in Solicitation

307
01:00:43.080 --> 01:00:56.809
Jim Donlon | NSF: institute awards are going to be in a form of cooperative agreements up to twenty million dollars, for up to five years, with an average yearly budget of around four million dollars do not submit total budgets outside the specified range, and the solicitation

308
01:00:56.820 --> 01:01:11.920
Jim Donlon | NSF: in addition, some specific ah, some special instructions pertain to the way budgets and team one are to be structured, as you heard in the ah alluded to in the Webinar um theme description, and we'll cover that a little bit more later.

309
01:01:12.210 --> 01:01:26.320
Jim Donlon | NSF: Solicitation does require that you plan for travel budgets for your personnel. This will be key to enabling effective multi-organizational synergy. So plan, according to what fits your proposal at a minimum. The solicitation requires that you plan for site visits and for pi meetings.

310
01:01:26.410 --> 01:01:44.590
Jim Donlon | NSF: Please also anticipate that institute operations require a significant investment in administration. The solicitation anticipates this with a requirement that a managing director or a project manager who is distinct from the leap I be provided for. In addition to this role you might wish to plan for supporting administrative roles of various times.

311
01:01:44.600 --> 01:01:50.339
Jim Donlon | NSF: What is appropriate depends on your proposal and the resources available to you at your local institutions.

312
01:01:50.410 --> 01:01:56.210
Jim Donlon | NSF: Be sure to account for these costs in your proposal, and demonstrate how this will facilitate your Institute's effectiveness.

313
01:01:57.190 --> 01:02:10.629
Jim Donlon | NSF: Now let's turn to the proposal steps and we'll start with the first major milestone in the preliminary proposal. Preliminary proposals are required as a condition for submitting a full proposal. The prelim consists mainly of a six-page project description.

314
01:02:11.480 --> 01:02:18.189
Jim Donlon | NSF: The description is intentionally very short, and within it we ask that you focus on the intended scope of the proposed research activities.

315
01:02:18.200 --> 01:02:28.620
Jim Donlon | NSF: Therefore we require these sections. Please read carefully the description of these sections and submission instructions, and what we're looking for. You see them listed here A. Through E.

316
01:02:28.790 --> 01:02:32.209
Jim Donlon | NSF: The section headers are mandatory and must be in that order

317
01:02:32.740 --> 01:02:48.930
Jim Donlon | NSF: the page limit and suggested, Ah! Section lengths are intended to keep the focus of your preliminary proposals on your intended research code and the suitability of the proposal personnel to carry out those activities. This is not the stage where we expect to review detailed plans for research and other activities.

318
01:02:49.140 --> 01:03:02.770
Jim Donlon | NSF: The submission guidelines also list sections not to include this includes The section results from Prior and a self-support and the project. Description as well as standard proposal. Sections, you see, excluded here on the right do not include those in a preliminary proposal. One

319
01:03:03.990 --> 01:03:13.620
Jim Donlon | NSF: again, we emphasize that the preliminary proposal is required, meaning a full proposal will not be accepted, if not following a preliminary proposal reviewed under the solicitation

320
01:03:13.890 --> 01:03:21.380
Jim Donlon | NSF: you will submit it to a theme, and the full proposal will be submitted to this by the same lead organization to the same theme.

321
01:03:21.830 --> 01:03:31.339
Jim Donlon | NSF: Preliminary proposals will be internally reviewed by program officers who are primarily looking for responsiveness and the likelihood that the concept is is going to be competitive.

322
01:03:31.350 --> 01:03:39.419
Jim Donlon | NSF: After just a few weeks we'll send an advisory determination by email to the P. I. And the authorized organizational represented listed in the proposal.

323
01:03:39.920 --> 01:03:45.259
Jim Donlon | NSF: In addition, narrative feedback may also be sent by email directly to the lead. Pi.

324
01:03:45.280 --> 01:04:01.530
Jim Donlon | NSF: This advisory type of determination indicates whether we encourage or discourage the submission of a full proposal. This is a non-binding determination, meaning even if you've received the determination of discourage, you may submit a full proposal that further develops the preliminary,

325
01:04:02.380 --> 01:04:17.520
Jim Donlon | NSF: and Nsf. Intends to provide these determination within a month's time to provide you this feedback as early as possible in your continuing proposal development. Our experience with the same process in the last round showed the stage to be beneficial in proposal, preparation, and review.

326
01:04:17.730 --> 01:04:21.589
Jim Donlon | NSF: The prelim stage allows you to propose early plans for Nsf.

327
01:04:21.600 --> 01:04:39.229
Jim Donlon | NSF: Builds in more time for proposal development, and allows proposals to adjust plans based on feedback received, whether that be toward improvement of the preliminary idea, or to avoid the effort of preparing proposals unlikely to succeed in a competition such as this, in which there are very few expected awards that are

328
01:04:39.240 --> 01:04:41.620
announced. You in the sole citation

329
01:04:43.870 --> 01:04:56.090
Jim Donlon | NSF: we want to re-emphasize that preliminary decisions are non-binding. In any case, the decision to prepare and submit a full proposal. Should be taken. Carefully. Please Note that the proposal i'm sorry that the program

330
01:04:56.100 --> 01:05:07.339
Jim Donlon | NSF: may provide discouraged determinations to help members of the research community to make informed decisions about whether expending effort preparing proposals Unlikely to succeed is the right thing to do,

331
01:05:07.350 --> 01:05:23.820
Jim Donlon | NSF: and to undertake that significant revision; or if those ah personnel might wish to allocate the reference to other proposals. So remember to consult the Pa. Pg. Guide, and some preliminary proposals. If you have any questions about the um about the rules around this

332
01:05:24.400 --> 01:05:30.169
Jim Donlon | NSF: other questions arise about how your plans may change between preliminary and the full proposal.

333
01:05:30.180 --> 01:05:46.169
Jim Donlon | NSF: The full proposal must be submitted by the same lead organization that must be submitted to the same theme. Aside from that is expected that a great many things might change when you submit the full proposal, including scope of activities, personnel organizations, collaborations, and

334
01:05:46.390 --> 01:05:57.429
Jim Donlon | NSF: mit Ctl and a required single copy document in the full proposal provides a place for you to identify significant changes since the preliminary proposal, and we'll look at that requirement more closely when we discuss the full submission. One

335
01:05:58.560 --> 01:06:04.690
Jim Donlon | NSF: Another question that might arise is whether an individual may be listed as senior personnel on a preliminary proposal,

336
01:06:04.700 --> 01:06:09.409
Jim Donlon | NSF: and then later listed as senior personnel on a full proposal from another submitter.

337
01:06:09.510 --> 01:06:13.630
Jim Donlon | NSF: That is a submission based on a different preliminary proposal.

338
01:06:13.640 --> 01:06:20.100
Jim Donlon | NSF: Yes, this is allowable as long as the individual listed is on at most one proposal at each stage.

339
01:06:22.110 --> 01:06:35.220
Jim Donlon | NSF: We now turn our attention briefly to the highlights of ah proposal. Preparation for full proposals. Here let me preface by reminding everyone that the proposal preparation in the guidelines in the Pappt apply

340
01:06:35.230 --> 01:06:45.710
Jim Donlon | NSF: the solicitation, addresses, special instructions, or deviations from that procedure. So be sure to consult the Phpg. Where the solicitation does not have such special instructions.

341
01:06:46.160 --> 01:06:56.949
Jim Donlon | NSF: Among those special instructions is this twenty five page project description. That's a key element of your proposal. And here, at a very high level, we're showing you the required sections of the project description.

342
01:06:56.960 --> 01:06:59.190
Jim Donlon | NSF: Please include these headings, as shown.

343
01:06:59.320 --> 01:07:12.720
Jim Donlon | NSF: Note that we ask you to address several desirata that we group together as the strategic impact of the Institute to address under the Pa. Ppg required section Broader impacts. This is a convenience for you to group them under the required heading.

344
01:07:13.350 --> 01:07:26.280
Jim Donlon | NSF: So we think this Ah! High-level organization is a sensible order for those major sections. It's up to you in this case how you order a number any of these headings, and you can add other sections as you see fit.

345
01:07:26.700 --> 01:07:40.149
Jim Donlon | NSF: In the solicitation you may include the required results of You'll see that we allow for you to include the results of prior and a supplementary document rather than squeeze it in into the project description.

346
01:07:40.330 --> 01:07:47.890
Jim Donlon | NSF: There are several other important instructions related to other required or optional supplementary and single copy. Documents. So please take a look at those

347
01:07:48.470 --> 01:07:57.180
Jim Donlon | NSF: as with all proposal, submission, instructions. Read carefully all the instructions in the papp, and the solicitation, so you submit a complete and compliant proposal,

348
01:07:59.860 --> 01:08:15.979
Jim Donlon | NSF: as mentioned in the discussion on preliminary proposals, the full proposal will require that you submit a single copy document to help you link your full proposal submission to your preliminary proposal and call program officers attention to the significant changes.

349
01:08:16.720 --> 01:08:28.470
Jim Donlon | NSF: Um! I put an asterisk there under, so on. Single copy to emphasize to that single copy, and in a set parlance means it's one that's seen only by an asset program officials and not by reviewers.

350
01:08:28.490 --> 01:08:42.160
Jim Donlon | NSF: This document is going to identify the elements you see here in terms of Here's how it links to a prelim, and what changes are. You know You want us to know about that you made since the review of your preliminary proposal.

351
01:08:42.380 --> 01:09:03.770
Jim Donlon | NSF: You'll see that this um summary significant changes is meant to be brief. Two hundred and fifty words. So you know. Here, again, the brevity is an indicator of what we're looking for. Naturally, the full proposal is going to be a significant development beyond what was in your preliminary here. We just wish for you to summarize at a very high level the changes that department, noticeably from what had originally been envisaged in your preliminary

352
01:09:07.439 --> 01:09:20.769
Jim Donlon | NSF: letters of collaboration, so they are allowed for the full proposal. These letters are most helpful for you to substantiate the broader range of partners you intend to engage in your institute, and are therefore appropriate for such things.

353
01:09:20.779 --> 01:09:39.399
Jim Donlon | NSF: They're not necessary from the organizations that are in your proposal as awardy and sub-worthy organizations, because their roles and commitments are clear in your proposal they're self- evident, including too many of those can distract from the ones that are truly um value, added collaborations beyond those funded personnel.

354
01:09:39.920 --> 01:09:59.069
Jim Donlon | NSF: We emphasize here that letters of collaboration are encouraged, but letters of support are not allowed. This distinction is covered in the Ppg. So the inclusion of endorsements of so, or statements of enthusiasm from the project for the project or its personnel from your letter writer that's going to make your letter non-compliant.

355
01:09:59.130 --> 01:10:03.930
Jim Donlon | NSF: The letter should also not contain a description of the partners organization and personnel.

356
01:10:03.980 --> 01:10:09.370
Jim Donlon | NSF: Finally, do not include letters from any of the funding organizations listed at the top of the proposal.

357
01:10:12.420 --> 01:10:27.119
Jim Donlon | NSF: Okay, in short, the Ppg provides this as a recommended format for a letter of collaboration. We do check for this and flag for your bowl submission. If you've got non-compliant letters, so stick to this, and they'll be both easy to collect and avoid compliance issues one hundred and fifty.

358
01:10:27.130 --> 01:10:37.890
Jim Donlon | NSF: Um. This letter to keep in mind and maybe inform your collaborators. So this is for proposal. Documentation only serves to demonstrate that you have a substantive potential to collaborate.

359
01:10:37.950 --> 01:10:47.980
Jim Donlon | NSF: Sometimes proposers have difficulty getting a letter writer to stick to this brief form, because the letter writer has the mistaken understanding that the letter could obligate them to do a thing.

360
01:10:48.040 --> 01:11:04.639
Jim Donlon | NSF: Ah! Over such concerns they might wish to state additional limitations on the letter of collaboration, specifying the nature of the agreement to collaborate. We can tolerate that if the statements are brief, and if they refer only to activity that's otherwise fully described in the proposal as required.

361
01:11:04.980 --> 01:11:17.319
Jim Donlon | NSF: If the letter describes plans and capabilities that are not otherwise found in your proposal in the proper place, such as project description, or even the facilities and equipment and other resources document. It'll be considered non-compliant

362
01:11:17.730 --> 01:11:25.670
Jim Donlon | NSF: if the letter contains statements from the letter writer regarding their organization and their capabilities, the letter will be non-compliant.

363
01:11:25.720 --> 01:11:29.790
Jim Donlon | NSF: So any And again, letters of support are not allowed.

364
01:11:32.340 --> 01:11:54.110
Jim Donlon | NSF: Okay on submission and eligibility. Here we'd emphasize some of the important submission and eligibility requirements in the solicitation. You can see them listed here on the slide. They're documented more fully in the full solicitation. So there are two types of institutions who can apply as lead organization. But other types of organizations can be similarities.

365
01:11:54.150 --> 01:12:10.469
Jim Donlon | NSF: Um. Where submission limits are concerned. You'll notice that it's two per organization at each stage, and that there's a limit on the number of proposals that you can participate on at senior personnel. It's one preliminary proposal, or one full proposal that is, at each stage One

366
01:12:10.480 --> 01:12:27.189
Jim Donlon | NSF: and Api's and copies are our senior personnel, but senior personnel is a broader definition that's included in the Pa. Pg: You should look that up. Basically, if you're a faculty, level person or equivalent in your organization, your senior personnel

367
01:12:29.480 --> 01:12:56.929
Jim Donlon | NSF: questions about the application of organizational submissions limits based on the uh uh campuses of the same university system. So here's the guidelines for that. If you happen to be from such a university system. There's a Pa. Pg. Faq. That addresses this eligibility Uh: a distinct organization for the purpose of uh separate proposals has these kinds of uh features here. These organizations tend to have their own duns. Number separate sponsor project offices

368
01:12:56.940 --> 01:13:08.890
Jim Donlon | NSF: uh, and therefore can submit up to two proposals. Uh, see the full text of this in the link on the slide and contact us. Have your Aor contact us. If there are further questions about it.

369
01:13:10.560 --> 01:13:20.989
Jim Donlon | NSF: Okay. And now we're ready for your questions, and we have plenty to start with. We've been answering some in the module. We'll take; others live if they still

370
01:13:21.000 --> 01:13:36.209
Jim Donlon | NSF: and exist. Continue to submit questions to the Q. And A. Module. We'll address as many as we can also want to make you aware again that we are going to post these webinar materials, including this recording to the program page within just a few days to this one hundred and fifty.

371
01:13:36.310 --> 01:13:50.000
Jim Donlon | NSF: When I post the slides, you'll also find that there are some bonus slides at the end here that address some other frequently asked questions. This will just be for your later reference. Just a Ah, a handy. Ah list of

372
01:13:50.080 --> 01:13:58.079
Jim Donlon | NSF: popular Faqs. So with that let's turn our attention to the Faq. To the Q. And A. Module, and I thank everyone for your attention.

373
01:13:59.120 --> 01:14:00.980
Jim Donlon | NSF: So, um

374
01:14:01.370 --> 01:14:11.770
Jim Donlon | NSF: colleagues, let me know if you have anything teed up you want to answer live. I know Sylvia has volunteered to address a few of them that are still pending. Sylvia,

375
01:14:16.730 --> 01:14:22.389
Sylvia Spengler: I have some of these I want. I would like the group to be clear about.

376
01:14:22.400 --> 01:14:23.929
Sylvia Spengler: So

377
01:14:24.740 --> 01:14:31.949
Sylvia Spengler: the classic. The classic question is particularly in themes, two and three

378
01:14:31.970 --> 01:14:33.370
Sylvia Spengler: um.

379
01:14:33.870 --> 01:14:45.240
Sylvia Spengler: Should a proposal be driven by a single application versus a single foundational advance put in multiple application areas.

380
01:14:45.710 --> 01:14:48.569
Sylvia Spengler: My answer is that

381
01:14:48.860 --> 01:15:08.580
Sylvia Spengler: your Institute proposal is an integrated handshake between use, inspired and foundational Ai, and the best way you and your team do that is the answer to that question. There is not a preferred way of doing it.

382
01:15:11.180 --> 01:15:29.460
Jim Donlon | NSF: Thank you. So you know um. And by the way, theme lease, if you see anything, pop into the q and a module that's a theme specific, please, just but in and and we'll address that. Meanwhile, let me just take a look at some of these other program-wide ones. We have one are all key personnel expected to be funded

383
01:15:29.470 --> 01:15:34.830
Jim Donlon | NSF: uh if unfunded, should they not be listed as unfunded collaborators. Okay,

384
01:15:35.020 --> 01:15:37.590
Sylvia Spengler: that's a free proposal.

385
01:15:37.600 --> 01:15:43.639
Sylvia Spengler: It's at the pre-proposal stage. I was going to say things change a lot between cup and lip

386
01:15:44.330 --> 01:15:45.290
Sylvia Spengler: okay,

387
01:15:45.300 --> 01:15:47.089
Sylvia Spengler: just put it in

388
01:15:47.470 --> 01:16:08.099
Jim Donlon | NSF: so. Um the a a key distinction. So nsf doesn't have a term key personnel per se, but they they've got the in the term senior personnel. Um, don't put any don't. Put any senior personnel in your in in your Budget submission. If they're not actually receiving budget. Okay, uh talk to us about any other questions you have about that.

389
01:16:08.280 --> 01:16:10.139
Jim Donlon | NSF: I hope that helps

390
01:16:10.270 --> 01:16:11.570
Jim Donlon | NSF: it's

391
01:16:14.870 --> 01:16:24.199
Jim Donlon | NSF: somebody's asked a question about the one-page supplementary document that can be used for results of prior

392
01:16:24.270 --> 01:16:30.829
Jim Donlon | NSF: prior Nsf. Support. Right? That's one page in total, not one page per personnel.

393
01:16:32.460 --> 01:16:33.700
Jim Donlon | NSF: It's

394
01:16:33.920 --> 01:16:45.970
Sylvia Spengler: are all key personnel expected to have active or past and F, and as a funds the Solicitation doesn't make any mention of that it's there's no restriction like that in the Solicitation

395
01:16:48.240 --> 01:17:04.309
Jim Donlon | NSF: our ffrdc is eligible submit the proposals. No, they're not an eligible organization. Type Institutes, fire, education and nonprofits are eligible organization types, but we do have maybe let's say it might even be in my slides here.

396
01:17:05.540 --> 01:17:25.460
Jim Donlon | NSF: No. So the the thing about F therdc's is Go to the papp and find the guidance on ah funding to Federal organizations, and you'll find that Nsf. Usually does not directly fund other Federal organizations. But if you include them as a funded organization, or you

397
01:17:25.470 --> 01:17:40.539
Jim Donlon | NSF: tend to, I recommend that you obtain one of these letters that's described in the Pa. Pg. Of the unique capabilities of the Ffrdc. And how those unique capabilities relate to your proposal.

398
01:17:40.630 --> 01:17:45.600
You have questions about doing that. Please email the program box, and we'll help you with it.

399
01:17:55.830 --> 01:17:58.100
Sylvia Spengler: The bailout of two of them.

400
01:18:03.100 --> 01:18:12.130
Jim Donlon | NSF: Well, the preliminary proposal be provided to the review panel, together with the full proposal. No. When we convene a review panel, they are there

401
01:18:12.140 --> 01:18:29.699
Jim Donlon | NSF: reviewing the proposal. Ah, in question, and in addition i'll reiterate that. Ah, they don't even see the single copy document in which you describe what your changes were since the prelim they're just Ah! Evaluating the proposal as it stands at that at that submission.

402
01:18:29.830 --> 01:18:45.929
Hector Munoz-Avila | NSF: Um gema. An additional point regard related to that. There was a question about whether you know the the panel will know that this proposal was not invited or was invited, and the answer is, no, the panel will not know that information.

403
01:18:45.940 --> 01:18:47.920
So it's like roly- the ice again.

404
01:18:48.470 --> 01:18:49.559
Jim Donlon | NSF: It yeah

405
01:18:49.570 --> 01:18:59.100
Jim Donlon | NSF: um. We have a theme to question, sonar. I don't know if you had a chance to um to notice that last one submitted. I'll give you a chance to read it if you like.

406
01:19:00.530 --> 01:19:03.989
Jim Donlon | NSF: Oh, okay, This just got submitted. It looks like,

407
01:19:06.030 --> 01:19:11.290
Jim Donlon | NSF: and I want to make sure to clear the way for any theme specific things. The

408
01:19:11.300 --> 01:19:14.780
Serdar Ogut: yeah, this is this is, you know what I

409
01:19:14.870 --> 01:19:20.290
Serdar Ogut: when I meant research tools or modes, you know

410
01:19:20.420 --> 01:19:37.349
Serdar Ogut: it could involve any one of those different modes There's not one preferred, or this combination is ideal. It's to address the problems and materials

411
01:19:37.360 --> 01:19:46.110
Serdar Ogut: advancing foundation. Why active? So there's no we wanted to keep it very agnostic in that sense.

412
01:19:49.450 --> 01:19:57.900
Jim Donlon | NSF: Good. Uh. If there's any, follow up to that just type it into the Q. And A. Module. And thank you, sir. Um, there's A. Let's see.

413
01:20:02.520 --> 01:20:16.859
Jim Donlon | NSF: So an early submission here in the Q. And A. Module. I'm sorry that we skipped past it. I wonder how the innovations and Ai in the Institutes would relate to, or be encouraged to converge with the Ai advances in industry.

414
01:20:17.290 --> 01:20:18.580
Jim Donlon | NSF: Um!

415
01:20:18.590 --> 01:20:41.619
Jim Donlon | NSF: So the solicitation doesn't say anything about that right? Um. Ah! Depending on your submission and what it is you're trying to do. It might be a good idea to describe that in as much as it pertains to your Institute strategy. But we don't have any kind of expected to, or encourage to language in the solicitation where we don't restrict things or encourage things.

416
01:20:41.630 --> 01:20:54.130
Jim Donlon | NSF: We don't have a We don't have a sort of um definitive answers on such questions. I hope I know that's a non-committal answer. But that gives you all the all the running field you need to write your proposal.

417
01:21:01.650 --> 01:21:02.290
Jim Donlon | NSF: Stay!

418
01:21:02.300 --> 01:21:08.769
Jim Donlon | NSF: Have some people offering to collaborate with one another. That's fun. Talked about that one

419
01:21:10.460 --> 01:21:11.920
Jim Donlon | NSF: about the

420
01:21:12.920 --> 01:21:17.289
Jim Donlon | NSF: and it's a results of Nsf. Support right one page total right.

421
01:21:17.440 --> 01:21:40.859
Jim Donlon | NSF: The rationale for this is we don't need twenty-five pages of this from you right, and you you, you weren't going to write twenty-five pages of it. If you had to squeeze it in the project description as the Ppg required. We're giving you a full page much more space than you would like to have allocated in your in your project. Description right to, uh, briefly, briefly, describe your results of prior an assessment.

422
01:21:44.520 --> 01:21:54.350
Jim Donlon | NSF: Are there best practices or lessons learned from the existing Ai Research institutes that can be shared with the scientific community and potential applicants. Yes,

423
01:21:54.360 --> 01:22:13.529
Jim Donlon | NSF: we have baked a lot of those lessons into this webinar places where we've been kind of reiterating and and and encouraging emphasis on certain aspects of the desirata, and talking about us aspired research, and how not only proposals, but institutes are evaluated on. How well they do this integrated practice.

424
01:22:13.540 --> 01:22:26.699
Jim Donlon | NSF: This is all lessons that are kind of confirmed in our experience with Institute so as much as possible. We've been trying to reflect that through both in the revisions to the solicitation and throughout the Webinar,

425
01:22:29.640 --> 01:22:34.629
Jim Donlon | NSF: should the order of the pre proposal subsections be followed strictly. Yes, the

426
01:22:38.350 --> 01:22:41.089
Jim Donlon | NSF: how many senior personnel are appropriate!

427
01:22:41.100 --> 01:22:56.100
Jim Donlon | NSF: There's no restriction to that. There is, of course, a restriction to the number of pis that can be listed just by virtue of the practicalities of the asset. Cover page that that cover page will allow for one of the Api and four other other copies,

428
01:22:56.330 --> 01:23:15.949
Jim Donlon | NSF: but you will list all of your senior personnel in a supplementary document as described in the solicitation Um, The number of senior personnel are enough so that you are covering all of the disciplines and activities and anticipated projects, full scope of what it is you intend to do at Institute scale.

429
01:23:16.330 --> 01:23:23.359
Jim Donlon | NSF: So sorry I know that's non-committal. It's. But it's better to give you open field than this to give you a number.

430
01:23:24.270 --> 01:23:26.190
Jim Donlon | NSF: Ah! What's up? Historical

431
01:23:26.200 --> 01:23:28.219
Jim Donlon | NSF: over subscription rate.

432
01:23:28.290 --> 01:23:57.739
Jim Donlon | NSF: Oh, you want to know about um about proposal pressure or funding rate? Unfortunately, we didn't really come prepared to talk about that. We only really came prepared the Webinar to talk about the current solicitation. Um. But just bear in mind that we identify specifically a number of awards anticipated in each theme, and where that number is one, or where that number is two. That's highly selective for the number of proposals sent in. This is part of the reason why we build in the preliminary proposal,

433
01:23:57.750 --> 01:24:12.400
Jim Donlon | NSF: so that you have an opportunity to describe to us briefly your prospective proposal, and we can give you feedback on that. And if you receive a discourage with remarks that you know really can't realistically be addressed and be competitive.

434
01:24:12.410 --> 01:24:35.620
Jim Donlon | NSF: That's our early sign to you uh about how you might economize that effort. That's the best we can do for you. Funding rates are very funny thing in a in a program where number of submissions is highly variable and expected awards is sometimes one. So think of it that way, and we've tried to give you all the conditions we can to help You make informed decisions about that hope that helps

435
01:24:39.450 --> 01:24:46.150
Jim Donlon | NSF: again. Theme leads. If you see something popping in here that needs answering, please just jump in!

436
01:24:47.660 --> 01:24:50.810
Jim Donlon | NSF: Was a Sarah Sharif, already addressed that one.

437
01:24:52.430 --> 01:24:53.589
Serdar Ogut: I'm looking at it.

438
01:24:53.600 --> 01:24:54.599
Jim Donlon | NSF: Okay?

439
01:25:01.890 --> 01:25:14.349
Jim Donlon | NSF: Why, you're reading the Sudar. Let me just take a look at in the pre-proposal page limit very tight. Yes, intentionally. Should we identify the role of every senior personnel listed?

440
01:25:14.360 --> 01:25:25.290
Jim Donlon | NSF: My best advice to you is read the proposal preparation instructions for preliminary proposals. It indicates to you what should be addressed, and even

441
01:25:25.300 --> 01:25:54.000
Jim Donlon | NSF: even pretty much signals The rationale for what we want you to address a really significant aspect of the preliminary proposals for us to see. Here's your intended research scope, and the personnel you have that are meant to deliver on that right to imagine. It's easy to write a preliminary proposal with all the right goals, but without the personnel to back that plan up. So you you want to describe

442
01:25:54.010 --> 01:25:57.810
Jim Donlon | NSF: that linkage to the extent that it helps you make that case

443
01:25:57.820 --> 01:25:59.240
Jim Donlon | NSF: that helps.

444
01:25:59.250 --> 01:26:09.459
Jim Donlon | NSF: Ah, so there um is asking. Can materials, research use for designing a hardware

445
01:26:09.930 --> 01:26:21.650
Serdar Ogut: design, or the theme means only I had to be used for materials. Design Ai to be used for materials. Design is certainly one aspect of it. I didn't quite understand

446
01:26:21.890 --> 01:26:30.300
Serdar Ogut: the other part materials research used for designing Ai hardware sign. Maybe,

447
01:26:30.470 --> 01:26:38.480
Serdar Ogut: Sahart, I'm: assuming this is safer. You can send me an email on that. I'll be happy to Okay,

448
01:26:40.620 --> 01:26:59.740
Jim Donlon | NSF: So here's a question: is the evaluation of the pre-proposal based on compliance with the call or is it based on the strength of what is proposed. I hope we've made clear both in this um recent answer to a related question, and in the Webinar that the preliminary proposal is evaluated for its um

449
01:26:59.750 --> 01:27:13.499
Jim Donlon | NSF: for its responsiveness to the solicitation, including the theme to which it's submitted, and our judgment of the likelihood of a full proposal based on that prelim to be competitive in the theme.

450
01:27:14.070 --> 01:27:25.240
Jim Donlon | NSF: So it's very much about the content of what you um of what you submitted there. Not just a sort of mechanical compliance. If you will hope that helps

451
01:27:26.960 --> 01:27:32.229
Jim Donlon | NSF: Hector, do you want to address theme? Three lives since I see you, you're toiling away there?

452
01:27:32.850 --> 01:27:39.170
Hector Munoz-Avila | NSF: It was a question about tool a string to the Ib: Very soon a specific Nsf:

453
01:27:39.440 --> 01:27:48.900
Hector Munoz-Avila | NSF: Yeah. In Cpu's teams, tool or tools and no it doesn't It doesn't have to be a specific to to exist in Nsfa Institutes.

454
01:27:51.390 --> 01:27:53.779
Hector Munoz-Avila | NSF: Thank you, Hector. I hope that helps

455
01:27:55.320 --> 01:28:08.050
Jim Donlon | NSF: somebody has a noted Thank you. Everybody for helping me to manage the Q. And a module here. Thanks for cleaning up. I see this. Ah! Recent question are a ipis required to be part of the proposal or

456
01:28:08.060 --> 01:28:20.979
Jim Donlon | NSF: two domain experts suffice. So let me say a few things, one I previously stated, and I will restate where the solicitation doesn't require or restrict a thing we don't require or restrict a thing.

457
01:28:21.800 --> 01:28:40.419
Jim Donlon | NSF: Having said that, please note the very, very serious goal of an Ai Institute to advance Ai in its knowledge and methods and understanding of artificial intelligence. So how you select personnel ought to reflect the full range of the six desirata and the solicitation.

458
01:28:46.400 --> 01:28:56.259
Jim Donlon | NSF: Okay, we're out of questions. I see like we can see, three o'clock is ticking down really fast. Last chance to type something into the Q. And A. Module.

459
01:28:59.290 --> 01:29:22.409
Jim Donlon | NSF: I see people dropping off because you have meetings here at the top of the hour. Thank you, everybody for joining us. I'll remind you we're going to post these materials to include a few. Ah Bootleg Faq at the end. Here two slides worth of other questions that sometimes come up. Please feel free to consult that. Get in touch with the program contacts with any questions. We're very happy that you joined us. We look forward to your proposals,

460
01:29:22.420 --> 01:29:23.599
Jim Donlon | NSF: bye, everybody.