By now, everyone has heard about the OpenAI-Navier-Stokes saga and the resulting existential crisis in mathematics. My own experience with the rapidly improving capabilities of AI in materials simulation has pushed me to ask the same question of my own field: will materials theorists become redundant? At the risk of spoiling the conclusion, my answer is no, because I have spent my entire career trying to make myself redundant and I haven’t quite succeeded. 2006: My first paper as a PhD student was the Li-Fe-P-O phase diagram. It took the better part of my first year, most of it spent debugging calculations, fixing input files by hand, and queuing multi-day jobs on a congested cluster. The analysis and writing took less than a month. 2011: Fast forward five years. I started pymatgen and custodian. Input generation and error recovery stopped being manual. The Materials Project launched the same year, with pymatgen and custodian as building blocks. Anyone could now pull up the phase diagram of any chemical system in seconds (the Phase Diagram app was the first app I built for MP, other than search). Even then, the first indications of an existential crisis for materials theorists were already there. Many experimental groups increasingly decided they could run their own calculations. Phase stability and battery voltage predictions no longer required a simulation specialist. So I moved on. I declined collaborations in which nothing interesting was asked of the theorist and the request amounted to routine DFT calculations, even when a high-profile paper was on offer. I turned my group to problems that were still hard: ion diffusion, luminescence in phosphors, defect properties. 2022: A decade later, universal machine learning interatomic potentials, or foundation potentials, arrived on the scene, with my group contributing our small part to their development. A whole swath of once-hard problems suddenly became easy: long-timescale, large-scale molecular dynamics, phonons, elastic constants. With MatCalc, we tried to make them easier still: a property calculation that once needed a bespoke workflow and days of DFT on a supercomputer became a few lines of Python and minutes on a laptop. Then LLMs matured. LLMs and foundation potentials, orchestrated via MatCalc, turn out to be an especially lethal combination. Even with a fairly high-level prompt, the LLMs are now able to generate a physically grounded sequence of calculations, write the glue code, and deliver the analysis in a nicely formatted report with graphics. Today, my group members and I can test a PhD's worth of hypotheses this way. More than ever, it seems like the end of the materials theorist is nigh. So why do I believe otherwise? My observations of users of these powerful new tools (fellow scientists, postdocs and students) showed me that the human driver of these tools matters more than ever. There is a certain aspect of - for lack of a better word - taste in choosing which questions are important and which hypotheses would probabilistically yield the most useful insights. A poorly posed initial question results in AI scientific slop, an unfortunate side effect that is now becoming far too common. Indeed, the works that I am personally most proud of from my group are not the highly cited tool papers, but those that leverage these tools to make an unexpected materials discovery (e.g., Zhenbin Wang's discovery and confirmation of the first Sr-Al-Li-O compound and its application as a phosphor) or gain new insights (e.g., Xiang-Guo Li's work on short-range order in the NbTaMoW "high-entropy" alloy and our collaboration with the Asta and Ritchie groups on dislocation motion in this alloy). The theorist whose value was "I can run the DFT calculation" started becoming redundant in 2011. The theorist whose value is knowing what to calculate now wields the throughput of an entire research group. To paraphrase a popular superhero, with that great power comes even greater responsibility to work on harder problems. That, after all, is what science has always been about for me. ** This post is about AI, but not AI-generated. Any dashes, semi-colons, etc. are my writing tics. AI was used to fix grammar and phrasing.**

We held our first monthly group lunch at the excellebt South and East restuarant at the NUS Shaw Foundation Alumni House today. It is also an opportunity for us to say farewell to Chebgrui Xu, who has done exrraordinary work with us for the past 2 months as an exchange student. Chengrui will be heading back Tsinghua University soon, but his work here will definitely continue. We wish him all the best on the remainder of his undergraduate studies, and hope to see him again in future!

Prof. Shyue Ping Ong has been awarded the National Research Foundation's Returning Singaporean Scientists (RSS) award. The award supports the AI Foundry project, which aims to develop an integrated discovery system that closes the loop between AI prediction, automated experimentation, and continuous learning. Initial programs target high-entropy alloys and ceramics for fusion energy and aerospace applications, with a modular design that supports expansion into semiconductors and clean energy in line with Singapore's RIE2030 priorities. The project builds on two decades of open infrastructure from Prof. Ong's group: founding development of the Materials Project, creation of pymatgen, development of the Materials Graph Library (MatGL), and the 2022 work on foundation potentials that demonstrated screening across millions of crystal structures. Read the full profile at the NUS College of Design and Engineering.

We took our very first Materialyze group photo at NUS today!

At Materialyze.AI Lab, we are on a mission to pioneer the integration of theory, experiments, and AI to accelerate the discovery and deployment of breakthrough materials. We welcome postdoctoral and PhD applicants. We believe that the right team can make all the difference. Our Core Values At Materialyze.AI Lab, our core values guide everything we do. Here are the key principles we stand by: Integrity : We practice integrity in all forms. We are honest and fair to fellow group members and collaborators. We have a zero-tolerance policy towards plagiarism and falsification of results. Excellence: We strive for excellence in everything that we do. We stand by the quality of our science. We aim to develop scientists with great analytical, technical and communication skills. Teamwork: We believe great teamwork is the key to great science. We share and discuss ideas freely. We strive to build great collaborations, both within and outside of the group. We contribute actively to the materials science community. Opportunities at Materialyze.AI Lab We welcome postdoctoral and PhD applicants with expertise in either theory & AI or experimental materials research, or ideally, a combination of both. Theory & AI in Materials Discovery Develop and apply machine learning and AI models (e.g., ML interatomic potentials, generative design, reinforcement learning) to predict and design novel materials. Perform first-principles and molecular dynamics simulations to model structural, thermodynamic, and electronic properties. Contribute to open-source software, benchmarks, and datasets that advance the global materials community. Experimental Materials & AI Integration Synthesize and process functional materials relevant to batteries, aerospace alloys, and semiconductors using solid-state, solution, or thin-film methods. Apply advanced characterization techniques (XRD, TEM, SEM, spectroscopy, electrochemistry, etc.) to probe structure–property relationships. Collaborate with theory and AI researchers to validate predictions, generate datasets, and develop high-throughput/automated experimental workflows. Experience in developing autonomous laboratory systems is a strong plus. Application Process Submit Your Application : Start by submitting a cover letter, CV, and 3+ referee contacts via this form . Initial Review : This usually takes place within two weeks of your submission. Initial interview via Zoom: We invite a subset of applicants for an initial interview with Prof Ong and his postdoctoral associates. In this initial interview, you will be requested to prepare a short 10-slide/10-min presentation summarizing your research experience, what you hope to get out of your time and what you feel you would be able to bring to our lab. You will be informed of the results of this initial interview within a week. Full interview for postdoctoral applicants: We invite you for a 2-hour interview with the whole group. You will be requested to give an hour-long presentation that outlines your research accomplishments and future plans in greater detail. Conditions permitting, this interview will take place in person and all travel expenses will be covered. In the event this is not possible, the interview will take place via Zoom. This interview is also an opportunity for you to get to know the group members and ask any questions you may have. Admission application for PhD candidates: You will be asked to submit an application for admission to the Doctor of Engineering (Materials Science and Engineering) program via the NUS application portal . In your application, please indicate in the section asking about your research interest and potential supervisor that you would like to work with Professor Shyue Ping Ong. Offer : If everything goes well, we will extend an offer to join our team. We will provide you with all the details regarding salary, benefits, and start date. What to Expect as a Team Member Once you join Materialyze.AI Lab, you will be welcomed into a supportive and engaging environment. Here are some things you can expect: Onboarding : We provide a comprehensive onboarding process to help you get settled. You will receive training and resources to ensure you are set up for success. Mentorship : You will have access to mentors who can guide you. They will provide support and advice as you navigate your new role. Continuous Learning : We believe in lifelong learning. You will have opportunities to attend workshops and conferences to enhance your skills. Team Building Activities : We organize regular team-building events to foster camaraderie. These activities help strengthen relationships and create a positive work culture. Join Us Today If you are excited about the prospect of working at Materialyze.AI Lab, we encourage you to apply. We are looking for individuals who are passionate, driven, and ready to make a difference.

Foundation potentials (FPs), i.e., universal machine learning interatomic potentials with periodic-table-wide coverage, are now proliferating. New pre-trained FPs seem to appear almost daily, making hyped-up claims about performance. For those not deep in the field, it’s becoming increasingly difficult to separate signal from noise. Here’s my expert, very opinionated take on how to choose an FP that actually fits your needs. 1️⃣ Dataset quality and size drive everything. Architecture is a secondary consideration. If you’re working with inorganic crystals or condensed matter, there are really only two datasets that matter today: OMat24 and MatPES (yes, I’m involved with the latter, but I’m being objective here). OMat24 wins on sheer size and broad coverage near equilibrium, but its convergence criteria based on MPRelax parameters are a bit loose for true PES quality, i.e., the kind that people use to train plain MLIPs. But at least OMat24 is constructed from static calculations, not merely sampled from intermediate steps in a relaxation trajectory or MD, such as MPTrj. MatPES is smaller, but its parameters are modernized for PES accuracy - the latest pseudopotentials, strict energy and force convergence, etc. Our experience is that the predictions are much more reliable, and it is much easier to work with. It is the only option if you want r2SCAN PES information, which for many systems, is far superior to PBE. Pick either one. The community should retire MPTrj — it was never meant for MLIP training and was just the best available dataset in the early days of FPs. 2️⃣ Don’t mix datasets unless you really know what you’re doing. I’m still amazed how often people blend OMat24, Alexandria, and MPTrj thinking it improves generality. It doesn’t. Mixing a high-quality dataset with a low-quality one yields a low-quality model. 3️⃣ Choose architecture efficiency over novelty. Unless you’re after new physics beyond PES quantities (like charges or magnetism), focus on architectures that give the best accuracy with the fewest parameters and lowest inference cost. At this point, the major FP architectures perform similarly in terms of accuracy. The real differentiator is computational efficiency. Some of the more models are more than 1000× slower  at inference and can barely handle systems beyond 100 atoms. This defeats the entire purpose of a FP: being a practical surrogate for DFT . If you just want to play around with different FPs to check their performance on a property that matters to you, MatCalc is a good choice to get this done with a minimal amount of coding.