American Physical Society
Task Force Plenary
2/22/24
[Standing by to begin]
[Standing by to begin]
>> Jan: Hello.
>> Ian: Hello, Jan, how are you?
>> Jan: Good. How are you doing?
>> Ian: I'm okay.
>> Jan: I apologize, I have to leave around the half hour. So, I'll do my best to catch up after that.
>> Ian: Okay. All right. I'll do that...
Okay. All right. Slides are uploaded as a PDF as well.
Joel's back. I'm not -- I didn't hear from everybody today. Give people another minute or so.
Okay. People can you see my screen?
>> Yes.
>> Ian: Okay. The one thing about this particular configuration is I don't easily see the chat window when the slides are in presentation mode. I do see when people raise their hands. So, people should do that when they want to interrupt. And I guess...
It's 3:05, let's get started. Okay. Thanks, everyone, for coming. Let's see... I had a quick bit of introduction and then I wanted to talk about -- there's two sections to talk about today. Last week we discussed communications and partnerships and technical working groups. Today we will discuss career development and DEI activities. That is the last of the major sections in the -- at least the outline to discuss. And sticking to the schedule with our normal schedule, I would hope that we would meet on March 7th. At that point, I hope we have a document that we can go through. So, after today we will have done the major sections and people will do some writing. Everyone is encouraged to contribute to the text in the document. There is a proposal on the membership section which I have seen a lot of people commenting on. And Joel and I will post the transcripts of the first two meetings and this meeting in a place that we tell you as soon as we -- I think it's probably gonna be on Indico. But probably for the meetings themselves. We will discuss that afterwards, and if it's not that, we will send the correction.
But the idea would be to finish this section -- to finish this exercise and then to -- for two weeks just have some writing and then to go through the idea of it being on the meeting on March 7th to verify there's not some big section we have missed and make sure that we can at some point converge on a report. So, I was gonna cover the career development. And then I hope Maria is gonna cover DEI activities. Though I -- I don't see all the participants -- yes, she's there. So, hopefully she can go through the bullet on that.
>> Maria: I don't know if I needed to make slides or if I can show from the document or if you did slides?
>> Ian: So, I think you can do whatever makes you comfortable. I'm happy to go through the document. But I've made a few slides for Career Development. So, we'll go through those and then go to DEI.
So, there are two questions from the primary charge about career development. One is should the CPSC create awards to recognize the work of Early Career and/or established scientists in HEP? Suggest the auspices under which such reports might be created. And the other is what mechanism should the CPSC use it promote discussion on career development, including recruitment, training, and retention, as well as potential for help categories to improve opportunity and sustainability for HEP? One is relatively simple, I think, and one is complicated. At least as I understood it from Petra, the awards were seen as generally a success. So, I believe this is a statement, people can choose to discuss it. But the purpose of the awards is to highlight the purpose of the awards in the contribution of science and computing, help people get a job and advance in the next job. That's the purpose of the awards as far as I'm concerned. And to whom? Postdoc? Junior faculty? All of the above? And multiple awards, what type? This is an open question, which we should decide, is there any value to working with industry to get a small financial award? Or is the recognition sufficient? Certainly on the CPAD side, it's recognition. And there would be annual call, and the CPSC chooses the number of winners, two per year? But that's up for discussion. The one thing I thought would be interesting to put in is an assessment which is a 5-year follow-up to see if the awards were making the desired impact. I could imagine giving awards to postdocs, those might be the people who were identified by industry for poaching. It would be interesting to assess if it helped with people's jobs or retention or both? And so, are there any thought -- I'm opening the floor for discussion. Any thoughts about awards? No one cares about awe wards.
>> I raised my hand.
>> Hands up.
>> Ian: Oh.
>> A lot of hands.
>> Ian: The things did not go to the top.
>> Joel: Okay, I'll go first. Yeah, I know how it arranged them. I just wanted to make one small point which is that, you know, we always think about awarding early career people. But I suppose I can speak up for a different age group but of computer scientists, not myself. There is an advantage in giving some awards to more senior people because it sets a target for -- it sets a target to aspire to.
>> Ian: Okay.
>> Joel: And these don't have to necessarily be financial awards. I mention this because there have been times where some of us have tried to put people who have essentially focused and made major contributions to computing in HEP as part of, you know, the general awards that are available through DPF. And those have never gotten through. And some of the cases were so evident and obvious to people like me, at least, that it's absolutely astounding. So, I think that maybe -- you understand what I'm saying.
>> Ian: Yep. So, the idea that perhaps the -- in the proposal -- we might have sort of three classes of awards. One would be a postdoc level, one early career, and one a distinguished service or more senior award?
>> Joel: For example, yes.
>> Ian: Okay. Liz.
>> Liz: Yeah, I was going to comment on the number of awards per year. I would guess -- but I could be argued out of it -- that it might be difficult to compare, say, a postdoc from the there are you community with one from cosmology or from high energy physics. So, I'm just thinking we might need more than two.
>> Ian: Great. You have been in management a long time. You're one of the primary responsibilities, generate an integer when asked. So, what is that number?
>> Liz: Well, I mean, I could -- in the experimental program I think, you know, we would go along -- neutrinos, cosmology, and -- and collider. And then maybe a theory. So, four? I don't know. That's just the random integer on when asked answer. I haven't begin it a lot of thought.
>> Ian: Topically-based on order of four.
>> Liz: Yes.
>> Ian: Peter.
>> Peter: Currently they can mark with the CV. I think some element of merit and career stage is also required. I like the idea of different types. I don't think reserving one for late career rather than just saying merit is probably better.
>> Ian: Okay.
>> Peter: The plan to work with industry to get small, financial awards. I think that if there are -- if there's interest in sponsorship, that's great. But our plans shouldn't be based on requiring sponsorship from industry for it to work, right? I think we have to be able to say recognition is sufficient and this is an esteemed merit. The possible disadvantage of getting industry involved is we're then exposing our best people to industry as part of the purpose.
>> Ian: I hadn't thought of that. Good point. I didn't mean to cut you off.
>> Peter: Sorry, no, no. That was all the points I had to make.
>> Ian: Jan.
>> Jan: Right. I was mainly going to make the same point that Joel raised. I think it's important to have some sort of continuity. Also to just have, you know, some sort of -- something to aspire to. Not something that, okay. You do this and then you're done. You change careers and move out. I don't think competing with industry is a problem. For postdocs, I wouldn't worry about them all leaving to industry because there's always going to be a fraction that just want to earn money and want to do good work and there's a fraction that want to do research. And drive their own agenda. We should encourage that in general. I think it's easier to assess that with junior faculty to see how they get, you know, since they're already on the track to see how they progress through those stages based on the awards that we're getting.
>> Ian: Tulika.
>> First, I support what Joel said, I think it would be important to have some -- at least one that goes to a senior person. With regards to the earlier sort of early career/postdoc/junior faculty, I'm okay if we want to increase the number. But perhaps not make it too many. And it would be good if the committee was given the guideline that they should spread the awards out into different areas or sub-areas instead of saying one for neutrino, or one for cosmology. I've known in some years CPAD has not gotten a whole lot of nominations. And in constraining them to choose per area might be a big constraint.
>> Ian: That's a good point. And I think there is a push to keep it small also keeps it prestigious and so, that helps. Giordon?
>> Giordon: Yeah, I have a lot of comments. But I'll try to make this concise. First, I think the comment about, you know, whether or not people leave to the industry because they want to make money versus, you know, doing something they like. I don't know if that's necessarily a fair characterization because there are many cases where even in academia it doesn't pay well enough to actually support being able to stay alive. And people with physical disabilities who are going to need a lot more support. And they just find it much easier to get that support outside of academia than within it. So, I don't know if that's necessarily going to be a reason -- a good characterization there.
The way I see these awards is that we should be supporting not only the work, but also trying to reduce the burden where possible. So, for example, if the awards were a little bit more focused towards underrepresented minorities. Or institutions that do not have as much financial support as other institutions. That would be a good way to prioritize it. Partially because it allows more people to be exposed to scientific computing than would not happen otherwise. You can see some, for example, at least in the school of physics and having people go over that to South Africa to teach software and computing for a week or two weeks and that tends to be a very well-received and a very successful program. We could do the same thing with these awards where we say, you know, there are existing programs in the US. I think Women Who Code and Kids who Code, these are programs dedicated to minority for software and computing. And even doing an award to allow those people from say, HEP, to join those programs and help contribute to that effort is a lot more beneficial than just awarding it to somebody at Stanford. Because they have plenty of computing resources already on some level. Not that there aren't people at Stanford who deserve these awards. But we should try to prioritize it a little bit better, or at least spread out the money a little bit more.
>> Ian: Two things, not sure that there's any money yet, unless we get it from industry. And the other is would you then propose that maybe a focused award on like -- an award on encouraging diversity or an award on -- so, it would be -- it would be an award for activities in the area of diversity, not simply technical areas?
>> Giordon: Yeah. I would like to see something like that.
>> Ian: Okay.
>> Giordon: My one concern is typically a lot of these awards tending to focused on -- and seem to be not very diverse or inclusive at all. And we start to see these sorts of, you know, racial biases. And just because the people involved with these computing technologies are not that diverse. And the awards are not being given out in a very diverse or representative way.
>> Ian: Okay. Other comments? Okay. I think I have enough notes that I -- that we could write something and people could then comment on them. All right. Let's go to the next part of this part. Which is this is the part that I thought was a little bit -- was more complicated. And so, what I wanted to do initially was to stop and talk a little bit about our goal for it had section. Which is about career development, recruiting, training and retention. So, I could imagine a variety of goals that as a committee we would like to see. One of them is are we trying to increase the visibility and prestige of software and computing contributions to allow software and sciences. Meaning that if you were concentrating on software and computing as your contribution to the program, but were in the physics department making sure that the community as a whole, including your local academic community, valued those contributions in the same way they values physics analysis? Are we trying to keep people in software and computing in science and out of industry? Are we trying to increase the number of stable positions at software and computing universities? As someone who used to be in the lab, the lab tended to be a more stable place for a software and computing than in software at a university? Are we trying to change that? Those I thought were goals. And I wanted to -- this next part I wanted to phrase in a way that was not blasphemous. But are there aspects of the status quo that are beneficial that we don't to want use? From my perspective, the fact that software and computing has a lot of turnover is also one the reasons why it moves quickly. Sort of there are new ideas that people go out into industry, sometimes they come back, sometimes they don't. But that if you look at -- if you -- like progress is made one going away party at the same time sort of model, there is something to be said for the fact that we turn people over also in software and computing. And in some sense, new ideas -- we don't want to sort of stifle innovation. So, comments? Maria?
>> Maria: Yeah, a couple of things. I don't think that people going to industry is a failure. For me as a mentor, if somebody wants a job in green tech and gets them, I consider that as a success. So, I wouldn't want to come off as this people that think that industry is garbage. So, that's cultural. Also because we don't have -- I think academia, sorry, I think permanent positions, one in four, one in five. So, we don't want to say, you know -- to make a bottleneck on the postdoc space or on the grad students. I don't think that would be productive to the training of the, you know, US workforce, basically. If we shut down -- if we close down on number of students and postdocs. I think what we would like to see, increase the number of stable positions -- more than software and computing, I would say hybrid positions in, you know, physics and computing. That's my first thing that I would say. We would like to make sure that those who want to stay in academia have an opportunity to do. So but without saying if you leave, you failed, or we failed.
>> Ian: Okay. Matthew?
>> Matthew: Hi. Sorry, excuse me. I'm a bit sick. Yeah. So, I -- I was a bit surprised to hear the position of people leaving for industry and then coming back be kind of put forward as a thing that might happen on a regular basis. I mean, I asked in chat and I have been given some examples of where this has happened. But I think that overall from my experience, I'm on the younger side of the field, but that's something that we don't see happen on a regular basis. And so, I don't necessarily view that as something that we can hold up as a normal thing that happens. Or that we can kind of view as there being a real exchange there. I think I'm -- in general, I'm not really -- I agree on -- I'm not really at all concerned about people leaving for industry in the sense that I think that's fine and great and that there's, you know, people from my kind of generation of Ph.D. students at CERN have gone on to do some really amazing things in industry that I know they're quite proud of. But I'm much more kind of concerned about the people that want to stay doing particle physics but who have put in substantial amounts of time in software and computing and then find that there is no positions for them.
And so, as Giordon also kind of mentioned, are leaving not because thing, oh, man, the big bucks of industry. That's much more interesting than science. But that there is realistically doing -- trying to find a N+1 postdoc is not a sustainable career path.
>> Ian: Yeah. One thing -- commenting on both of your comments together, actually, I think one of the things that we maybe want to talk about as a goal is I agree with you that up to now it has sort of been a one-way door that you leave for industry and then you don't come back. But one -- I think at least locally here at Flatiron, one of the thing we have been discussing is as part of our DEI programs, trying to find ways to have people come back from wherever they were and be contributors to academia. There are lots of people who had to make choices about -- whether their level of support, or they left academia for reasons not because entirely they wanted to. And opening up those paths to of legal coming back actually can diversify your workforce too. And so, it seem -- like it's been kind of -- it's been a difficult -- I agree that it hasn't happened. But I'm not sure that's a good thing.
>> Matthew: I would say Flatiron is doing awesome work and I'm happy to hear that. I think moving forward it's good to try to promote these ideas. But yeah, I think unfortunately Flatiron is kind of a shining example there of rather than the standard.
>> Ian: Tulika.
>> Tulika: In my view, we're trying to do both, create a technically adept workforce for our field as well as creating one that can help our country, so to speak, in all areas. Industry and otherwise. Now, there is -- I see students come in with different levels of preparation. And in here there are DEI issues as well. I think some of the programs that now both NSF and DOE are promoting are useful to get people maybe on a slightly more equal footing. And so, I think CPSC could sort of play a role in actually trying to get these different efforts together in one place. Maybe in that annual workshop or whatever that might be held.
And so, that's one role I see CPSC doing. Right now there are various university groups, there are things like SEP. That maybe we have a central place. And I guess second is to sort of increase the importance of sort of training and how that is valued by our community in S & C. We are responsible for doing it. But traditionally. Our field would be a lot of emphasis on, you know, students have electronics and hardware experience. Can we now start to give the same kind of acknowledgment to training in software and compute something and somehow enhancing that would be useful as well. But it's more difficult.
>> Ian: It's more difficult. But in some sense, it's also very impactful. Or could be. Charles?
>> Charles: So, I think one of the most important things we can do is get the various institutes to recognize the value of someone with a mixed background as a mix between software and computing and physics. I can think of more than just a hand full of postdocs who have left the field because they could not get a job in academia because they were very interested in the software and computing and the physics. And that aspect of their career development was just not recognized.
So, I don't know how we do this. Especially without money. How do we get institutes to realize that, you know -- well, to support people who have a lot of computing background and who are really going to be spending a lot of their time doing computing as opposed to just physics? And if we can get the institutes to actually value that and recognize that, then I think that would be a large fraction of our job.
>> Ian: I agree. It was actually -- probably on the next slide. But I think -- I need you to talk next about sort of how we might do that. Because I agree that it's an important -- it is one of the most important things that we can do. It also -- I'm struggling to find out -- to think about how we might do that. Because my thought well, maybe if you could encourage joint appointments. You don't want them between physics and computer science. It's not -- what's valued in computer science is not necessarily what we value in software and computing as much. Or we see it as -- you might end up giving people like setting them up for failure as they weren't successful in either department as much as they would like to be. So, the question is how to sort of make a focus on the application of software and computing in addition to the main science and still have that be seen as recognized as valuable. And yeah. We'll talk about that next.
>> Charles: A lot of what we do is not computer science per say. It's software engineering, but not recognized as computer science.
>> Ian: Right, it's not academic computer science in the same way. Verena.
>> Verena: I wanted to say something that's almost the same as you, Charles. I think this prestige is really important. I have come across at least, I don't know if others would agree several, you know, very talented young scientists who have been given the advice: Go work on a hardware project if you want a faculty position. And there's sort of this idea you don't get job it is you work in software and computing. I've heard from several places. I would like to think that maybe things are changing. But I don't know if others have sort of come across this. And maybe having -- I think having like data like information about like that indeed, you know, and places are hired with kind of more of this profile and that they can -- there are lots of funding opportunities as well as things like that. I also have heard that some folks feel that in some of these comparative review panels and things like that, it's not always clear that software and computing is as valued as other tasks.
So, maybe sort of trying to yeah, gather -- I don't know how we could do it. But maybe, for example, when I find myself in these kind of conversations, it's difficult to sort of have -- if we had like numbers or some arguments, you know what I'm saying? That could be useful. I don't know if that -- if that would be helpful. But just brainstorming here. It's a difficult problem.
>> Ian: I think it's a problem that needs in some sense an approach across the board. Because I think the experiments -- at least in the experimental program, they don't help this because the people they put forward as the best postdocs are leaving. There's not as much of a recognition that this wouldn't have been done without this software in computing. And at the same time during the start of LHC, there was a point where basically an entire like entire funding agency said, like, now is the time you switch to doing analysis. We ended up losing large members of the software and computing side. At the agency level, at the post-institute level and also at the community level and the experiments, that we need to figure out how to make this point. Peter?
>> Peter: Yeah, I would like to -- somewhat echoing the same point. But I would like to emphasize modern science and most frameworks is increasingly computational. It's computational science. Compute asking a tool of the trade. If you want to do anything in fluid dynamics that isn't a spherical or whatever situation, you're ending up doing numerical simulation, for example. That's distinct from computer science, which is, you know, often human-computer interaction, compilers, whatever. Is distinct from software engineering in many ways. And you would like people to recognize that computing and science can be a tool of the trade just the way maths is for a theorist or, you know, and in fact theorists nowadays, the name of the game is in many aspects doing symbolic algebra. You get the computer to do your algebra for you. It's a tool that if you can't use it, you're unskilled, right? And I think the purpose of the awards is to raise the profile of this as one of the skills that a physicist should have. Not to mark people out as being distinct from physics as such.
>> Ian: Right. Yeah. Though in some sense I wonder if the -- if the awards then certainly -- I sort of agree with the point of this is something that should not be seen as distinct from physics. But -- and the application of physics. And I wonder if like -- while I think the awards are valuable, I hope we're not further making a separation that shouldn't be there.
>> Peter: I don't think the idea of the awards, but cautioning that the discussion here is talking about it as a different activity to physics. And, you know, we need to -- we need to see the point of view that -- that we aren't awarding computer scientists with this, we're marking physicists being particularly skilled in the tool of the trade.
>> Ian: Telling someone they're really good at mathematics, that's not seen as a bad thing. Joel? You're muted. Joel?
>> Joel: Sorry. There were two points that I want to make. The first one is that as I am asked to write letters for positions these days, I noticed that more and more positions are kind of cross-disciplinary. Where physics is seen as a -- an important use case and testing ground and maybe even a breeding ground for ideas for machine learning, for example. So, there are faculty positions that are now being posted that involve presence in two departments. Or presence in one department, but with a strong affiliation with another.
And if these wind up producing successful career paths, they could in part answer some of the issues that we're discussing. And maybe we should try to advocate for such positions. I mean, I think there's a lot more emphasis off now in the community for sure in learning how to I feel like train -- train machine learning procedures to handle things that you used to do for a lot of straight and massive coding. So, this I think is an opportunity to try to number one communicate with a segment of the computer science community. And number two, to advocate for people to have positions where they have a presence in both the analysis and computing camps and high energy physics, and the computer science camps and the universities. Now, other people probably know a lot more about this and I hope they'll comment. But the second point I want to make is that there is always been a missing component of the way our agencies think about things. And there have been times where There have been rays of home where it looks like this is gonna change. Specifically to do any physical sciences you need computing. And there is computing is part of what I like to think of as a -- as a national research infrastructure in the physical sciences. And there is no visible means really of supporting this activity even though we all recognize that it's critical.
Something like this organization could vociferously and aggressively promote that idea. NSF has occasionally come close to recognizing this. But they hate long-term support for their initiatives. They like to get them to succeed in their original context. And then they like to pass them off to someone else. But I think we desperately need this for long-term computing support. If we could promote the idea that the academic community in high energy physics and the lab community are responsible for producing -- for producing and supporting this national research infrastructure and it has value beyond HEP, then maybe we can get more respect for these efforts in the universities and more career paths. Which in the end is the basic solution to this problem. If we can do that, then a lot of the things that we're discussing will cease to be problems.
So, that's my 2 cents and I hope somebody will comment on either or both of those.
>> Ian: So, on the first one, Joel, the idea would be a push forward jab that is data science or computational science tracks might be a solution. ML is sort of a little -- is the flavor of the month. But data science has -- might have a longer path than that?
>> Joel: So, these tend to be joint positions. You have a foot in the physics community and in the university and you're doing physics analysis and whatever else is implied by being in that physics community. And you are contributing through your activities and that to computer science. So, you know, there are specific people, for example, that see that there are -- that they want to have HEP use cases to try out their different machine learning packages. And, you know, our use cases are somewhat unique for a number of reasons. Which I think you all appreciate and I won't go into. If we can continue -- if that continues to be the case, then I think there's room -- there will be joint interdisciplinary appointments in universities. I think this is a theme that is -- is important in the universities and it's accelerating, it's increasing. So, that's what I want to say about that. I don't think it's really sending physicists to computing positions in the universities. Okay?
>> Ian: Okay. Maria.
>> Maria: Yeah, I wanted to elaborate a little bit on what Joel was saying. Because of being recruited for one such position. And basically, what's happening in the last I'm gonna say three, four years is that, you know, a decade ago there was a big push in computer science that was mainly driven by this. The statisticians, the mathematicians and the pure computer scientists. Now there is a bit of back pedaling there. In the sense that people who are pure data scientists are not very useful broadly in the following way. They come in, they look at the dataset. Give advice on which algorithm to use. But that they don't particularly care about the problem. You know? In my group we have the interns from the data science program at Stanford. And, you know, they run their algorithm. The algorithm gives the answer. They don't know what the answer is and they don't care -- they really don't want to know what it does.
This is -- I'm saying this bluntly. It's not a criticism, by the way.
And so, a variety of universities have started slightly different programs which is really dual -- dual training. So, computational physics, computational biology, you know? At Stanford we have in addition to the data science department, we have computational mechanical engineering. For example, et cetera. So, those positions are quite useful. Like I said, a few universities are setting up these MS-level programs. And they're doing recruiting in this way. People with like data science and domain science appointments or training, et cetera.
And that's also a really important avenue for outreach. This is mostly not, you know, the top 20 universities in America is like 50 to a hundred that are doing this. Because it's very common that people with, you know, a background, first generation students, et cetera. Their parents don't let them go into pure physics or pure biology, they want to be able to get a job. So, this is a really good way for people to say, and tell their parents, I'm going to get a job because I'm computational something. It's kind of a back door into physics.
>> Ian: Tulika, they are drilling through my wall so I had to mute.
>> Tulika: Firstly, a comment on the last few comments. Definitely universities I see are hiring more people who sort of have a cross-disciplinary background. However, I feel that while that generally works for people whose expertise is in the area of machine learning and data science, it's less so in with people who might have, say, expertise in what I would call, you know, the non-ML-type. So, the core software and computing. So, somehow it would be nice if CPSC were able to achieve that right balance and so that -- for everything. From awards to these sort of positions that we encourage, you know, a little bit more of diversity and not just ML. Or whatever the national initiatives turn out to be. The other comment I have is following up on this, you know, chat thread regarding, you know, why hardware had been considered to be associated with physics.
So, why is software and computing, you know, struggling a little bit? And I think times are changing as software and computing become absolutely critical for our experiments. Just the way hardware is. But another important thing universities consider when they're thinking of hiring is what kind of funding the person will bring. And traditionally hardware has been associated with the big bucks. And so, people think, okay. If we hire this junior faculty, they'll be coming in with various big hardware grants. And again, things are changing there. But perhaps, again, the change is happening for things like ML and other such national initiative associated-projects.
So, if CPSC could have a role -- this is what CPAD has done so some extent in coordination with the funding agencies, they could have a basic research needs kind of study which really looks into these are important, critical areas for software computing to be put in. And that enables the DOE to put out a FOIA to say we are looking for people to work on these core software and computing topics. So, those were my thoughts.
>> Ian: I agree. It's interesting -- the interesting idea that the hardware is motivated by building up the large lab. Maybe ways that we can emphasize that software and computing also requires an infrastructure. Whether it's people or computers that involves reasonable startups, et cetera. Okay. In the interest of keeping a little bit of time for the last topic, what I wanted to do was -- this is actually just -- this is a summary from before the meeting on sort of what we had talked about. In terms of the mechanisms, we talked about awards. In terms of training, there's the option, as Tulika mentioned, we can also talk about workshops. Topical training. I also added the concept of maybe put a training for career guidance.
One of the things that I thought -- given that this is a committee that doesn't have any funding, one of the things it can do is prepare reports and one of the things that we might do is make guidance for the charge that there will be reports prepared on sort of the state of the field professionally in terms of what needs to change. And because -- in the very first -- defining the problem and where you would like to approach it is something that is sort of the first step. I wasn't sure -- if there were other obvious mechanisms that I didn't have on the slide, people should point them out now so they can go into the final report. That said, let us change gears in the remaining 15 minutes or so. And Maria was going to handle the DEI. So, I'm going to stop sharing.
Stop share.
>> Maria: Thank you. I think this is the right window. I took notes directing the document because this -- especially last time we had a good start on the DEI conversation. So, and I want to just share those. Can you see my screen? Maybe make it a little wider.
>> Ian: Very good.
>> Maria
>> Good. So, I took evidence from my experience is that computer software seems even less diverse than the rest of HEP. But we don't know for sure. Verena pointed out this. We should recommend a survey to the field to the new panelists as it's established. And not just for diversity, but like how many people do you have? How many of your postdocs are computational? Et cetera, et cetera. Something that became really clear during the pandemic is that computer and software has a lot of the potential being accessible in many different ways.
One is it doesn't require expensive experiment and can be done at any university. It's kind of the flipside of the conversation we were having now. Hardware people bring big bucks. Yeah, but software people are cheaper. Both students and faculty, right? I don't need $2 million for an electronic microscope. Just need stuff they already have. It can be done virtually. And so, people who have small children or elderly parents or people with mobility challenges or other disabilities, you know, are less affected by the in-person aspect of it. You know, my experiment is underground in South Dakota. I can not go there. I can not go underground. I have been once. But I need more people to go with me due to my mobility challenges than the work I could actually do. So, that's an opportunity. And then recently think about, the comment I just made earlier, that people with dual training tend to have non-traditional backgrounds or career paths. You know, first generation, immigrant background, low-income. And not maybe in HEP, but I have a broad -- what was the word? I have a number of colleagues that came back into academia from industry using this path. Machine learning people especially. I can think of like three or four people just at Slack that were recruited to the machine learning initiative from industry in the last five, six years. So, that's -- that's -- so, those are all reasons why computing and software should be accessible for than other fields of HEP. But maybe it isn't yet.
So, that's the preamble that I would like to put in this document. Then there is a conversation that happened on Zoom which was crucial. Tulika reminded us. That, you know, the traineeship programs that joint university students and labs require US citizenship. And that is even for those programs that are especially written for DEI. So, I have one of those renewed grants from DOE. And it's, you know, our partner universities are California State colleges or community colleges with a lot of immigrant population and with a lot -- with some immigrant students on a green card or even people that don't have a visa.
So, that's -- that's a little bit of a contradiction. It is actually broader here because there is, you know, country sponsor of terrorism. And, you know, people who are born in Iran cannot get an account at NERSC. And that's absurd given the contribution of Iranian nationals to the US HEP program over the years. That's pretty bad.
So, this is something we should flag for the DOE. Clearly, you know, we can write it in this document. But that's something that they really should flag for DOE. That this is -- we're doing things that are really counterproductive here if the goal is to increase diversity.
Then there is a -- should I stop here? I'm sending it -- I'm saying too many things. Maybe I can go through the last one and then we look at comments. The last one is Tulika, do we want a sub-committee for DEI? Verena suggested that, you know, these member who is participate in the selection of panel members, et cetera. I'm a little nervous about this for the following reason. If you scroll through this document, all the sections have a bunch of volunteers. Both for writing and for reviewing. The DEI section has three volunteers. Me, disabled woman, Giordon, disabled man, Tulika, a woman with an immigrant background like me. So, I think if we do a DEI panel, it's gonna be the work of the usual suspects and it makes me really nervous. First of all, if you're on this call, please sign up for this section because these concerns, even if they're not minorities, it concerns the whole field. And I worry that if we do a sub-committee specifically, it's gonna be like, oh, this is the women's problem or the immigrant's problem or the disable people's problem. And I don't want to do that.
You know, when Giordon and I go to events like this one or travel, it's a nightmare for me to participate in, you know, for us to participate in the life of the community. And I don't think we should have to do all the DEI work on top of it. But that's just me. And I'm happy to hear -- yep. I see hands.
>> Ian: I'm gonna let you call out hands.
>> Maria: Ian, your land is up.
>> Ian: Three things. I think that machine learning is changing a little bit, the cost model of software and computing in ways that it will be seen as expensive and valuable to the university communities and not just that software and computing can be done anywhere. I'm working with the State of New York right now and they are proposing half a billion to build a resource to do just AI. And so, ML is changing the economics a bit. In ways that are probably good in the sense that it raises the profile and may make it more complicated in the sense that it will reduce the number of places you can do these things.
And then the other thing I wanted to mention was that -- so, the funding agencies are not the only places to get resources for this kind of work. For instance, the foundation gave money to Iris HEP which was used to support remote internships during the pandemic for places -- for people from lots of places. None of them in the US. So, while the US government has lots of rules about who they can support, not everybody is the US government. So, we should approach places with this in mind.
We can flag the DOE for the general sanctions. But we are -- we will struggle with that. And I'm happy to put it in the report as an issue. But it's -- we're having -- we even have the same issue which is basically that any place that's under the general restrictions, it's very hard to do anything. So, we can mention that it's a problem. But my guess is we will be mention -- that's the most we will do.
And I put myself down as a contributor to this section already. So -- or just now during your talk. So, all right.
>> Maria: Thank you. Who is next? Giordon.
>> Giordon: Yes. Can you hear me okay?
>> Maria: Yes.
>> Giordon: So, I think there's a lot I could say here. The one thing that I think maybe that might be missing or maybe you don't think of first thing is sort of the climate change impact of some of these soft wear computing activities. You know? A lot of times people will say a new HPC uses a lot more energy because the algorithms are producing a lot more impactful, for example, in the environment. What you typically see is that impacts those particular underrepresented computers a lot more than some of us who live in slightly better areas on some level. So, I think some of the DEI activities here need to also do the same sort of considerations that we also do with these new proposed colliders. That they are efficient and are taking into account the footprint that -- their carbon footprint as well. You know? It's one thing to say that we will make the field more diverse. But 20 years from now, if we kind of kill off all the communities who are the product of diversity because they can't survive climate change, then you kind of -- then you have a losing game at that point.
>> Maria: This is an excellent point, thank you. This is it? Do you have more, Giordon? Okay. And one down, two little comments from you.
>> Tulika: Thanks, Maria, for bringing up this point. That DEI should be the responsibility of the entire committee. I hope that's the way we can make it and do it that way. As long as the committee is following-up on DEI and it's not left over, I think this is great. You mentioned some good points about how in principle software and computing should be more accessible to people who have other responsibilities. And it should have a better -- let's say demographics in terms of diversity. However, we also said it doesn't. So, is our intuitive, you know -- look, when I look around in my own experimental collaboration. So, then the question to ask is why not? And are there any concerns with regards to the other kind of climate. The climate within the software and computing sort of field? Are we welcoming to people who are -- sorry -- are we welcoming to people who are coming, you know, from different areas? So, I just wanted to raise that.
>> Ian: Also, I'm looking for the button. But I wanted to agree with Tulika. But I wanted to suggest that this is -- this particular area is part of also the pipeline. And I think one of the things that we're seeing around here is that the computational skill level of the people who come in as computational scientists even at Flatiron is not as high as we would like it to be. That there are excellent domain scientists and their computing skills are weak. Not all of them. But some. And one of the things that we're finding there's a real need for training. And I think this also goes down to the level of DEI is that if the people who had opportunities to do -- take an extra class in computer science or to be exposed to computing early, that this is a place that we can certainly close the gap because I think everyone has made this point that it should be more accessible -- more accessible and therefore better represented than it is. But I think it's a problem that -- that -- yeah. This is a place that we could also help with in terms of training. And I think throughout the entire field, certainly we would help them like us. Tulika, your hand is back up?
>> Tulika: Yeah, I just wanted to jump on that. I want to do agree, but what bothers if me and what was reported to me was about the training programs and how they restrict to US residency. This requirement just takes away half of the Ph.D. physics program in Wisconsin-Madison. And so, you don't really have a pool to that students. So, we are constantly faced by non-US students who want to do software and computing, and yet the funds that we have cannot be applied to them. And so -- and so, it also actually creates another problem. It creates like two classes of students. Students who can take advantage of these nice traineeships and get a move and then, you know, get nice jobs perhaps earlier. And students who can't. It's creating a problem too. I know we can't do much with the DOE here, because this goes at the highest levels of government, but I think pointing this issue out, how we could do this better is just maybe some ammunition that they need.
>> Ian: And look at alternative sources of funding. The foundation has a stated goal of increasing the number of Ph.D. in hard science by about 10%. 10% representation and underrepresented minorities in the next decade. And to do that requires -- that requires also contributing -- it's money, too. And so, it's not money here at our local offices, it's contributed out to the universities to sort of at the training at the pipeline level. There are places working on this. It just might not be the government.
I'm gonna call --
>> Maria: With the Science Foundation, for example, we have this postdoc that we're doing a tie-back. And when we get the DACA students, it comes from other assignments.
>> Ian: Giordon.
>> Giordon: So, I think maybe the most important point that you've made, Maria, is going to be sort of the remote way of software computing compared to sort of other parts of the field of physics on some level. I guess historically if you wanted to do particle physics, you had to be in proximity to these larger experiments. But with software and computing, trying to maybe emphasize -- I don't know if "requirement" is a great word to say, but encourage remote friendliness a lot more. Especially hiring for software and computing rule would be a big plus. You don't have to deal with the overhead of relocating people right now. They're maybe living where they're living because they have to take care of their parents or someone in the household, something like that. And it makes it easier, on some level, to recruit a more diverse workforce here. Even though it doesn't actually happen for some reason.
>> Maria: It has started since the pandemic, I must say, this we have -- at Slack we hire some remote-only positions. And this is mostly for people with dual careers. So, yeah. Thank you for saying that.
>> I can treat a migraine --
>> Maria: Yeah. And the other beat here is NASA is doing this quite well. I didn't put it in my bullet points. But I want to add it. Support for public data and public software. So, I worked on an experiment with NASA. And all of our data is -- let me finish the sentence. All the data from the Fermi telescope is public, the data and the background model. And we do trainings worldwide. Including middle income and lower income countries. And so, having public data makes the field more accessible. Worldwide.
>> Ian: I think there's a theme that's coming out which should definitely be in the report about sort of the -- the dichotomy between what should be a very inclusive part of the field and what it is. And so, an understanding as to why that is and how it might be improved I think is -- would be a valuable thing for this part of the report to cover. Because there are lots of places in the field that requires travel to exotic places or requires expensive machinery. Software and computing does not as much. Are there other comments on this section? Okay.
>> Maria: People are dropping off so I'm gonna take a screenshot of this before we close because I didn't have time. Oh, maybe I can copy from the chat. Yes, I can copy from the chat.
>> Ian: So, it being 4:07 and there are people at CERN who it's getting late, what I would suggest is thank you for the great discussion today. I'm hoping that this turns into text into the report. Peter Boyle is going to the next meeting, which is great, it's 4:00. And we will -- we will reconvene in two weeks which is whenever -- March 7th. March 7th. And in the intervening time, everyone should try to commit text to the various sections and then we will have a discussion about it at that time. Thank you all.
>> Matthew: Thank you, bye.
>> Thanks, bye.
>> Bye, bye.