Maxim Ziatdinov
@MaximZiatdinov
Followers
2K
Following
85K
Media
707
Statuses
7K
I use machine learning to boldly go where no human has gone before in experimental sciences. Opinions are my own, but please don’t hold them against me.
Pacific Northwest
Joined June 2021
“I believe that a scientist looking at nonscientific problems is just as dumb as the next guy.” ― Richard P. Feynman.
2
11
98
This week, the Nobel Committee set a terrible precedent by awarding a Nobel Prize to someone with a relatively low h-index. What kind of message does it send to young researchers?!.
107
204
3K
A grad student should automatically receive a Ph.D. degree if they publish N first-author papers (N varies by discipline, but generally N>=2). In today’s competitive job market, you shouldn’t waste time writing a PhD thesis no one will ever read.
154
262
3K
What #LK99 has demonstrated is that the scientific community is perfectly capable of doing a peer review via arXiv and social media tools, openly and transparently, further underscoring how irrelevant traditional journal publications have become.
74
242
2K
I frequently come across bios, which are like, “I got my PhD from this fancy place,” “did my postdoc in that very famous lab,” “published many papers and won grants and awards,”… and zero mentions of what that person has really accomplished.
37
35
568
Students and postdocs should be allowed to submit and publish papers without their PIs. If your PI didn’t do any actual work (measurements, calculations, etc.), they shouldn’t be on the author list of your paper. #AcademicTwitter #AcademicChatter.
42
32
534
It’s pretty clear that the current model of science and scientific funding is a dead-end.
18
63
241
Suppose I want to add my dog as a co-author on my paper to justify using 'we' in an otherwise solo-authored paper. What should I list as his affiliation?.
45
14
236
In 20 years, no one will likely care about the paper you are writing/submitting right now.
43
14
204
@ItaiYanai Perhaps worth noting that some of the most culturally impactful movies, from the Shawshank Redemption to Blade Runner, never won Oscar. I think the same is true for many scientific papers.
2
4
209
I find science PR to be very nauseating. Every paper has to be presented as a paradigm shift that is going to enable revolution in materials, energy, blah blah blah. No one really believes in it, especially not papers authors.
6
15
201
@BenjaminDEKR Perhaps, a dead end? Imagine coming to work on AGI to change the world and ending up working on a digital waifu and re-creating Google's demo from 7 years ago.
4
6
200
@AlecStapp @Jason He said the same thing when campaigning in 2015, but his (1st) administration actions were quite the opposite
When foreigners attend our great colleges & want to stay in the U.S., they should not be thrown out of our country.
3
4
157
If your ML model is not grounded in physics, you'll be getting garbage outputs the moment you move away from the training data distribution. No amount of GPUs is going to help you.
7
13
152
Seeing non-academic Twitter discussing whether flat bands are a good indicator of superconductivity gives me hope for humanity.
5
8
148
Looking back at the papers I authored or co-authored over the past 10 years, I can identify 3-4 papers that I think ended up being *somewhat* useful. The rest was a tremendous waste of time.
15
2
136
LinkedIn: I’m happy to announce … blah-blah-blah… standard template. Twitter: We will try reproducing the experiment in that superconductivity paper and will be live-tweeting our progress.
3
7
119
What if, and please hear me out, we cease the habit of labeling every research paper as revolutionary and acknowledge that science is inherently a methodical and gradual process, one that values incremental advancements? It's time to abandon sensationalism.
11
15
111
What are the cool applications of AI for materials in industry? I see plenty of AI/ML for drug discovery but not so much for materials.
18
7
111
@Noahpinion Barbie, Oppenheimer, and Dune would disagree. No projector at home can match the Imax experience of watching Dune 2.
13
1
101
Academic publishing industry.
3
14
93
My frustration with a lot of research in condensed matter and materials is that it looks like nothing gets done beyond (often overhyped) publications. People chase one “sexy” topic after another, promising better or new types of devices, but it’s all BS (and they know it).
7
8
98
Materials science is a truly interdisciplinary field, where in order to be successful, you need to have a solid background in physics & chemistry, a hands-on experience with multiple characterization & simulation techniques, and a familiarity with AI/ML methods.
1
11
91
A periodic reminder that you aren’t supposed to “attach” yourself to a paper by your student or postdoc if you haven’t done any actual work.
8
7
80
The rate at which my Google Scholar citations grow seems to be somewhat decelerating. Therefore, from now on, every time I review someone's paper, I'm going to ask them to cite a minimum of 10 (instead of usual 1-2) of my papers.
12
1
83
I believe it is the sacred duty of a scientist to call out BS, regardless of the rank of the people involved. We must call it out in meetings, we must call it out on social media, and we must never, under any circumstance, cease from doing so.
8
9
82
I think we should all agree that including middle-author publications in your CV doesn’t make sense.
79
5
77
Do I understand correctly that once your citations stop growing (relative to previous years), you should quit doing science and start doing something else? At least, that’s my plan. Quit while you’re at or near the top of your game.
24
2
79
At this stage of my career, I feel absolutely zero benefit from going to conferences. How about you?.
42
1
75
I just realized that I co-authored several papers on deep learning with this year’s Nobel Laureate in Chemistry, lol.
7
3
78
It's never easy but sometimes you just have to close all the tabs in your browser and never look back.
5
6
71
I am reviewing a paper and it is really good and there’s not much to comment on. I already read it twice and still can’t find anything to pick up on. Crazy….
10
1
70
Journals should advertise upcoming papers as movie trailers. “From the creators of XX method…”, “An unprecedented development in area YY…”, “This December… One of the most awaited articles in …”, “Only in [the name of a journal].”.
8
7
65
Our work on the active learning of structure-property relationships in ferroelectric materials via the automated experiment in scanning probe microscopy was just published in Nature Machine Intelligence:
5
11
71
@kennymxu There are so many questions about gender and sex that I start wondering what exactly they do there. .
0
2
68
A fully Bayesian implementation of Heteroskedastic GP is now available in GPax:. 🔍 Handle varying noise in input data.🛠 Dual GPs for signal & noise modeling .🧠 Custom kernels & priors for deep customization. Notebook with example ->
3
13
72
Exciting news! Our paper on the in-house built AtomAI software package has been published in Nature Machine Intelligence. 1/6.
1
9
68
In case you weren't paying attention Gaussian process is now Artificial Intelligence.
5
6
72
I just told chatGPT to interview me about my recent paper and then write a ‘news article’ about it. And it did. My mind is officially blown 🤯
2
4
63
Most people go to conferences to i) make new connections, ii) drink with their buddies whom they haven’t seen in a while, and iii) do some sightseeing. Virtual conferences took all that away and forced us to listen to your boring science. Glad that in-person conferences are back.
3
2
60
If you spent time around people conducting sophisticated experiments that push the boundaries of knowledge, or those developing instrumentation for those experiments, you would realize that there is no way AI, in its current state, can replace them.
6
9
59
It occurred to me that a lot of scientific research these days is actually engineering (e.g., finding optimal material systems for specific applications), which often operates under the guise of science to bypass rigorous scrutiny of its deliverables.
4
3
54
In AI4Science, much of what is marketed as discovery is often merely optimization within the boundaries of existing theories. True discovery, however, requires transcending those theories—uncovering conditions where they break down. AI, trained on data produced using existing.
1
5
55
@SwipeWright Yes, the "educated vs. non-educated" polls are implicitly rooted in the belief that college graduates can think for themselves and are less susceptible to propaganda, but this is may not be the case.
3
1
56
My advice to grad students considering doing a postdoc:. 1) Don’t do it. Try finding a well-paying job in the industry instead. 2) If (1) isn’t an option due to e.g. visa restrictions, try to find a postdoc position in a national lab.
11
2
52
Never thought I'd see the Northern Lights right from my backyard 🤯
1
1
55
When my wife and I were dating, I liked to talk about my research, and she always pretended she was interested. But I didn’t realize that she was pretending at the time. I didn’t even know it for years later. But a few months ago, she told me this. And now I love her even more ❤️.
1
2
45
Me: The scientific publishing is broken, journals impact factors are meaningless, there’s almost no correlation with the research quality, blah-blah-blah. Me after my paper gets accepted in a journal with IF>30: Well, maybe it’s not that bad. It looks like the system still works.
3
2
52
“Funding for scientific research has become risk-averse, favoring established researchers pursuing incremental discoveries”
6
10
49
New AI tool just dropped. According to unconfirmed reports, it outperforms GPT-4 on arithmetic tasks. Big if true 🤯
3
6
47
My hot take - or, perhaps, a bitter lesson - is that not every science field needs to be AI-powered. If your experiment produces only about 50 data points per month, your design space is relatively small (or the ways in which you can explore it are limited), and you already.
4
4
51
I feel really bad for young folks entering academic research these days as there’s so much emphasis on writing bullshit papers and so little on learning to build things that last.
4
2
48
@paulg I expect it will also become a convenient scapegoat for the failures that had nothing to do with it.
2
0
44
@juliaioffe Welcome to the USSR. It also means that someone reported her for “антисоветчина”, just like in the good old days….
2
3
41
I firmly believe that right now in the US we need more engineers than basic science researchers.
12
4
43
Thrilled to share that our team at Oak Ridge National Lab & University of Tennessee has been honored with the R&D 100 award for the work on “Physics-Informed, Active Learning–Driven Autonomous Microscopy for Science Discovery!”.
5
0
45
Speaking of “self-driving” labs. This is a typical setup for many surface science experiments (from my postdoc days). Good luck automating the whole thing. It is made for humans, not robots. But what you can do is use AI to automate data collection, analysis, & sample navigation.
4
2
42
Much of today's research seems to be a treadmill: incessantly writing papers to obtain more funding just to write additional papers, all while yielding little practical value.
4
4
43
Some academic email signatures (chair of this, editor of that) are longer than Daenerys Targaryen introduction.
5
2
44
Imagine if all the time and energy people spend on writing proposals asking for money to produce more papers, and the time and energy spent on preparing flashy PowerPoint slides, were channeled into actually building things.
7
7
42
@avt_im This guy has some sort of vendetta with everything “Bayesian” and seems incapable of explaining his point without insults.
1
0
40
AI/ML in a nutshell: if you have a sufficient amount of data you can train an ML model to make reasonable predictions on new inputs that are similar to training examples for pretty much any domain. It will also fail on new data outside the training data distribution.
6
1
40
Why are so many people working on using ML in computational sciences and so few on using ML in experiments?.
14
0
42
Deep neural network vs. linear regression for like 95% of the problems
0
7
39
If you don't make your codes available via GitHub/GitLab, I don't trust your research.
1
1
40
One of the biggest challenges in research today, especially for younger folks, is separating what has genuinely been achieved from what is merely promoted as a success via exaggerated research highlights. Because most of the latter is BS.
3
5
43
I was tricked into studying solid-state physics by being told that I would be able to move individual atoms and use this power to build new materials and devices. What about you?.
12
1
39
@Noahpinion Explelling all the nazis and getting normie tech bros into MAGA would be a true 11-D chess.
2
0
41
Kornia is an excellent library for (differentiable) computer vision that can work directly with Pytorch tensors. If you do deep learning in Pytorch I highly recommend using Kornia instead of torchvision.
0.5.4 release #kornia:.- Canny edge detector.- Patch-based augmentation (as in ).- morphology revamp -- now compatible with scipy.- Check the release notes: Happy coding!.#computervision #opensource #deeplearning
2
13
38
I’m not a theory person, but the part about the selection of U in the DFT+U has always given me cringes.
7
6
40
We love to say that we need more physics in ML in part because it justifies our funding/existence. However, my somewhat subjective observation has been that, beyond simple toy problems, physics constraints tend to do more harm than good to ML efficiency.
4
4
40
Reviewing for journals is essentially a consultation that scientists provide to journals and editors free of charge. This is quite remarkable, considering the substantial fees that other professionals, such as doctors and lawyers, charge for their consultations.
5
3
41
Open-access journals are a scam. Just submit to a standard journal and put a copy on arxiv. There's your open access.
3
5
39
Forget cloud. Forget HPC. All AI4Science will be done on $2k consumer GPUs.
2
4
39
True story: I was being ratioed into oblivion for telling the truth about middle-author publications. And when all hope seemed lost, @timgill924,the king of sociology, retweeted me. Soon,the twitter army of light came to save me. You can’t defeat us 'cos we have truth on our side.
6
1
38
You either quit academia while you still can or stay there long enough to see yourself become the Reviewer 2.
5
1
37
If you work on ML for science, trying to compete directly with places like DeepMind is silly. Better identify areas where DeepMind can’t shine (yet) because they don’t have access to the necessary equipment and expertise, such as advanced experimentation w/ neutrons,photons,nano.
5
4
38
When one is very knowledgeable in their field, they tend to quickly see everything that could go wrong and become resistant to new, high-risk, high-reward ideas. That’s precisely why we need generalists alongside specialists: to balance caution with creativity.
1
2
33
A few common red flags to watch out for:. - Referring to npj journals as Nature journals 🚩.- Calling Scientific Reports Nature Scientific Reports🚩. Personally, I believe all publishing should be done via arXiv. However, in my experience, researchers who make these.
5
1
39