More stories

  • in

    To excel at engineering design, generative AI must learn to innovate, study finds

    ChatGPT and other deep generative models are proving to be uncanny mimics. These AI supermodels can churn out poems, finish symphonies, and create new videos and images by automatically learning from millions of examples of previous works. These enormously powerful and versatile tools excel at generating new content that resembles everything they’ve seen before.

    But as MIT engineers say in a new study, similarity isn’t enough if you want to truly innovate in engineering tasks.

    “Deep generative models (DGMs) are very promising, but also inherently flawed,” says study author Lyle Regenwetter, a mechanical engineering graduate student at MIT. “The objective of these models is to mimic a dataset. But as engineers and designers, we often don’t want to create a design that’s already out there.”

    He and his colleagues make the case that if mechanical engineers want help from AI to generate novel ideas and designs, they will have to first refocus those models beyond “statistical similarity.”

    “The performance of a lot of these models is explicitly tied to how statistically similar a generated sample is to what the model has already seen,” says co-author Faez Ahmed, assistant professor of mechanical engineering at MIT. “But in design, being different could be important if you want to innovate.”

    In their study, Ahmed and Regenwetter reveal the pitfalls of deep generative models when they are tasked with solving engineering design problems. In a case study of bicycle frame design, the team shows that these models end up generating new frames that mimic previous designs but falter on engineering performance and requirements.

    When the researchers presented the same bicycle frame problem to DGMs that they specifically designed with engineering-focused objectives, rather than only statistical similarity, these models produced more innovative, higher-performing frames.

    The team’s results show that similarity-focused AI models don’t quite translate when applied to engineering problems. But, as the researchers also highlight in their study, with some careful planning of task-appropriate metrics, AI models could be an effective design “co-pilot.”

    “This is about how AI can help engineers be better and faster at creating innovative products,” Ahmed says. “To do that, we have to first understand the requirements. This is one step in that direction.”

    The team’s new study appeared recently online, and will be in the December print edition of the journal Computer Aided Design. The research is a collaboration between computer scientists at MIT-IBM Watson AI Lab and mechanical engineers in MIT’s DeCoDe Lab. The study’s co-authors include Akash Srivastava and Dan Gutreund at the MIT-IBM Watson AI Lab.

    Framing a problem

    As Ahmed and Regenwetter write, DGMs are “powerful learners, boasting unparalleled ability” to process huge amounts of data. DGM is a broad term for any machine-learning model that is trained to learn distribution of data and then use that to generate new, statistically similar content. The enormously popular ChatGPT is one type of deep generative model known as a large language model, or LLM, which incorporates natural language processing capabilities into the model to enable the app to generate realistic imagery and speech in response to conversational queries. Other popular models for image generation include DALL-E and Stable Diffusion.

    Because of their ability to learn from data and generate realistic samples, DGMs have been increasingly applied in multiple engineering domains. Designers have used deep generative models to draft new aircraft frames, metamaterial designs, and optimal geometries for bridges and cars. But for the most part, the models have mimicked existing designs, without improving the performance on existing designs.

    “Designers who are working with DGMs are sort of missing this cherry on top, which is adjusting the model’s training objective to focus on the design requirements,” Regenwetter says. “So, people end up generating designs that are very similar to the dataset.”

    In the new study, he outlines the main pitfalls in applying DGMs to engineering tasks, and shows that the fundamental objective of standard DGMs does not take into account specific design requirements. To illustrate this, the team invokes a simple case of bicycle frame design and demonstrates that problems can crop up as early as the initial learning phase. As a model learns from thousands of existing bike frames of various sizes and shapes, it might consider two frames of similar dimensions to have similar performance, when in fact a small disconnect in one frame — too small to register as a significant difference in statistical similarity metrics — makes the frame much weaker than the other, visually similar frame.

    Beyond “vanilla”
    An animation depicting transformations across common bicycle designs. Credit: Courtesy of the researchers

    The researchers carried the bicycle example forward to see what designs a DGM would actually generate after having learned from existing designs. They first tested a conventional “vanilla” generative adversarial network, or GAN — a model that has widely been used in image and text synthesis, and is tuned simply to generate statistically similar content. They trained the model on a dataset of thousands of bicycle frames, including commercially manufactured designs and less conventional, one-off frames designed by hobbyists.

    Once the model learned from the data, the researchers asked it to generate hundreds of new bike frames. The model produced realistic designs that resembled existing frames. But none of the designs showed significant improvement in performance, and some were even a bit inferior, with heavier, less structurally sound frames.

    The team then carried out the same test with two other DGMs that were specifically designed for engineering tasks. The first model is one that Ahmed previously developed to generate high-performing airfoil designs. He built this model to prioritize statistical similarity as well as functional performance. When applied to the bike frame task, this model generated realistic designs that also were lighter and stronger than existing designs. But it also produced physically “invalid” frames, with components that didn’t quite fit or overlapped in physically impossible ways.

    “We saw designs that were significantly better than the dataset, but also designs that were geometrically incompatible because the model wasn’t focused on meeting design constraints,” Regenwetter says.

    The last model the team tested was one that Regenwetter built to generate new geometric structures. This model was designed with the same priorities as the previous models, with the added ingredient of design constraints, and prioritizing physically viable frames, for instance, with no disconnections or overlapping bars. This last model produced the highest-performing designs, that were also physically feasible.

    “We found that when a model goes beyond statistical similarity, it can come up with designs that are better than the ones that are already out there,” Ahmed says. “It’s a proof of what AI can do, if it is explicitly trained on a design task.”

    For instance, if DGMs can be built with other priorities, such as performance, design constraints, and novelty, Ahmed foresees “numerous engineering fields, such as molecular design and civil infrastructure, would greatly benefit. By shedding light on the potential pitfalls of relying solely on statistical similarity, we hope to inspire new pathways and strategies in generative AI applications outside multimedia.” More

  • in

    AI copilot enhances human precision for safer aviation

    Imagine you’re in an airplane with two pilots, one human and one computer. Both have their “hands” on the controllers, but they’re always looking out for different things. If they’re both paying attention to the same thing, the human gets to steer. But if the human gets distracted or misses something, the computer quickly takes over.

    Meet the Air-Guardian, a system developed by researchers at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL). As modern pilots grapple with an onslaught of information from multiple monitors, especially during critical moments, Air-Guardian acts as a proactive copilot; a partnership between human and machine, rooted in understanding attention.

    But how does it determine attention, exactly? For humans, it uses eye-tracking, and for the neural system, it relies on something called “saliency maps,” which pinpoint where attention is directed. The maps serve as visual guides highlighting key regions within an image, aiding in grasping and deciphering the behavior of intricate algorithms. Air-Guardian identifies early signs of potential risks through these attention markers, instead of only intervening during safety breaches like traditional autopilot systems. 

    The broader implications of this system reach beyond aviation. Similar cooperative control mechanisms could one day be used in cars, drones, and a wider spectrum of robotics.

    “An exciting feature of our method is its differentiability,” says MIT CSAIL postdoc Lianhao Yin, a lead author on a new paper about Air-Guardian. “Our cooperative layer and the entire end-to-end process can be trained. We specifically chose the causal continuous-depth neural network model because of its dynamic features in mapping attention. Another unique aspect is adaptability. The Air-Guardian system isn’t rigid; it can be adjusted based on the situation’s demands, ensuring a balanced partnership between human and machine.”

    In field tests, both the pilot and the system made decisions based on the same raw images when navigating to the target waypoint. Air-Guardian’s success was gauged based on the cumulative rewards earned during flight and shorter path to the waypoint. The guardian reduced the risk level of flights and increased the success rate of navigating to target points. 

    “This system represents the innovative approach of human-centric AI-enabled aviation,” adds Ramin Hasani, MIT CSAIL research affiliate and inventor of liquid neural networks. “Our use of liquid neural networks provides a dynamic, adaptive approach, ensuring that the AI doesn’t merely replace human judgment but complements it, leading to enhanced safety and collaboration in the skies.”

    The true strength of Air-Guardian is its foundational technology. Using an optimization-based cooperative layer using visual attention from humans and machine, and liquid closed-form continuous-time neural networks (CfC) known for its prowess in deciphering cause-and-effect relationships, it analyzes incoming images for vital information. Complementing this is the VisualBackProp algorithm, which identifies the system’s focal points within an image, ensuring clear understanding of its attention maps. 

    For future mass adoption, there’s a need to refine the human-machine interface. Feedback suggests an indicator, like a bar, might be more intuitive to signify when the guardian system takes control.

    Air-Guardian heralds a new age of safer skies, offering a reliable safety net for those moments when human attention wavers.

    “The Air-Guardian system highlights the synergy between human expertise and machine learning, furthering the objective of using machine learning to augment pilots in challenging scenarios and reduce operational errors,” says Daniela Rus, the Andrew (1956) and Erna Viterbi Professor of Electrical Engineering and Computer Science at MIT, director of CSAIL, and senior author on the paper.”One of the most interesting outcomes of using a visual attention metric in this work is the potential for allowing earlier interventions and greater interpretability by human pilots,” says Stephanie Gil, assistant professor of computer science at Harvard University, who was not involved in the work. “This showcases a great example of how AI can be used to work with a human, lowering the barrier for achieving trust by using natural communication mechanisms between the human and the AI system.”

    This research was partially funded by the U.S. Air Force (USAF) Research Laboratory, the USAF Artificial Intelligence Accelerator, the Boeing Co., and the Office of Naval Research. The findings don’t necessarily reflect the views of the U.S. government or the USAF. More

  • in

    A more effective experimental design for engineering a cell into a new state

    A strategy for cellular reprogramming involves using targeted genetic interventions to engineer a cell into a new state. The technique holds great promise in immunotherapy, for instance, where researchers could reprogram a patient’s T-cells so they are more potent cancer killers. Someday, the approach could also help identify life-saving cancer treatments or regenerative therapies that repair disease-ravaged organs.

    But the human body has about 20,000 genes, and a genetic perturbation could be on a combination of genes or on any of the over 1,000 transcription factors that regulate the genes. Because the search space is vast and genetic experiments are costly, scientists often struggle to find the ideal perturbation for their particular application.   

    Researchers from MIT and Harvard University developed a new, computational approach that can efficiently identify optimal genetic perturbations based on a much smaller number of experiments than traditional methods.

    Their algorithmic technique leverages the cause-and-effect relationship between factors in a complex system, such as genome regulation, to prioritize the best intervention in each round of sequential experiments.

    The researchers conducted a rigorous theoretical analysis to determine that their technique did, indeed, identify optimal interventions. With that theoretical framework in place, they applied the algorithms to real biological data designed to mimic a cellular reprogramming experiment. Their algorithms were the most efficient and effective.

    “Too often, large-scale experiments are designed empirically. A careful causal framework for sequential experimentation may allow identifying optimal interventions with fewer trials, thereby reducing experimental costs,” says co-senior author Caroline Uhler, a professor in the Department of Electrical Engineering and Computer Science (EECS) who is also co-director of the Eric and Wendy Schmidt Center at the Broad Institute of MIT and Harvard, and a researcher at MIT’s Laboratory for Information and Decision Systems (LIDS) and Institute for Data, Systems and Society (IDSS).

    Joining Uhler on the paper, which appears today in Nature Machine Intelligence, are lead author Jiaqi Zhang, a graduate student and Eric and Wendy Schmidt Center Fellow; co-senior author Themistoklis P. Sapsis, professor of mechanical and ocean engineering at MIT and a member of IDSS; and others at Harvard and MIT.

    Active learning

    When scientists try to design an effective intervention for a complex system, like in cellular reprogramming, they often perform experiments sequentially. Such settings are ideally suited for the use of a machine-learning approach called active learning. Data samples are collected and used to learn a model of the system that incorporates the knowledge gathered so far. From this model, an acquisition function is designed — an equation that evaluates all potential interventions and picks the best one to test in the next trial.

    This process is repeated until an optimal intervention is identified (or resources to fund subsequent experiments run out).

    “While there are several generic acquisition functions to sequentially design experiments, these are not effective for problems of such complexity, leading to very slow convergence,” Sapsis explains.

    Acquisition functions typically consider correlation between factors, such as which genes are co-expressed. But focusing only on correlation ignores the regulatory relationships or causal structure of the system. For instance, a genetic intervention can only affect the expression of downstream genes, but a correlation-based approach would not be able to distinguish between genes that are upstream or downstream.

    “You can learn some of this causal knowledge from the data and use that to design an intervention more efficiently,” Zhang explains.

    The MIT and Harvard researchers leveraged this underlying causal structure for their technique. First, they carefully constructed an algorithm so it can only learn models of the system that account for causal relationships.

    Then the researchers designed the acquisition function so it automatically evaluates interventions using information on these causal relationships. They crafted this function so it prioritizes the most informative interventions, meaning those most likely to lead to the optimal intervention in subsequent experiments.

    “By considering causal models instead of correlation-based models, we can already rule out certain interventions. Then, whenever you get new data, you can learn a more accurate causal model and thereby further shrink the space of interventions,” Uhler explains.

    This smaller search space, coupled with the acquisition function’s special focus on the most informative interventions, is what makes their approach so efficient.

    The researchers further improved their acquisition function using a technique known as output weighting, inspired by the study of extreme events in complex systems. This method carefully emphasizes interventions that are likely to be closer to the optimal intervention.

    “Essentially, we view an optimal intervention as an ‘extreme event’ within the space of all possible, suboptimal interventions and use some of the ideas we have developed for these problems,” Sapsis says.    

    Enhanced efficiency

    They tested their algorithms using real biological data in a simulated cellular reprogramming experiment. For this test, they sought a genetic perturbation that would result in a desired shift in average gene expression. Their acquisition functions consistently identified better interventions than baseline methods through every step in the multi-stage experiment.

    “If you cut the experiment off at any stage, ours would still be more efficient than the baselines. This means you could run fewer experiments and get the same or better results,” Zhang says.

    The researchers are currently working with experimentalists to apply their technique toward cellular reprogramming in the lab.

    Their approach could also be applied to problems outside genomics, such as identifying optimal prices for consumer products or enabling optimal feedback control in fluid mechanics applications.

    In the future, they plan to enhance their technique for optimizations beyond those that seek to match a desired mean. In addition, their method assumes that scientists already understand the causal relationships in their system, but future work could explore how to use AI to learn that information, as well.

    This work was funded, in part, by the Office of Naval Research, the MIT-IBM Watson AI Lab, the MIT J-Clinic for Machine Learning and Health, the Eric and Wendy Schmidt Center at the Broad Institute, a Simons Investigator Award, the Air Force Office of Scientific Research, and a National Science Foundation Graduate Fellowship. More

  • in

    From physics to generative AI: An AI model for advanced pattern generation

    Generative AI, which is currently riding a crest of popular discourse, promises a world where the simple transforms into the complex — where a simple distribution evolves into intricate patterns of images, sounds, or text, rendering the artificial startlingly real. 

    The realms of imagination no longer remain as mere abstractions, as researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have brought an innovative AI model to life. Their new technology integrates two seemingly unrelated physical laws that underpin the best-performing generative models to date: diffusion, which typically illustrates the random motion of elements, like heat permeating a room or a gas expanding into space, and Poisson Flow, which draws on the principles governing the activity of electric charges.

    This harmonious blend has resulted in superior performance in generating new images, outpacing existing state-of-the-art models. Since its inception, the “Poisson Flow Generative Model ++” (PFGM++) has found potential applications in various fields, from antibody and RNA sequence generation to audio production and graph generation.

    The model can generate complex patterns, like creating realistic images or mimicking real-world processes. PFGM++ builds off of PFGM, the team’s work from the prior year. PFGM takes inspiration from the means behind the mathematical equation known as the “Poisson” equation, and then applies it to the data the model tries to learn from. To do this, the team used a clever trick: They added an extra dimension to their model’s “space,” kind of like going from a 2D sketch to a 3D model. This extra dimension gives more room for maneuvering, places the data in a larger context, and helps one approach the data from all directions when generating new samples. 

    “PFGM++ is an example of the kinds of AI advances that can be driven through interdisciplinary collaborations between physicists and computer scientists,” says Jesse Thaler, theoretical particle physicist in MIT’s Laboratory for Nuclear Science’s Center for Theoretical Physics and director of the National Science Foundation’s AI Institute for Artificial Intelligence and Fundamental Interactions (NSF AI IAIFI), who was not involved in the work. “In recent years, AI-based generative models have yielded numerous eye-popping results, from photorealistic images to lucid streams of text. Remarkably, some of the most powerful generative models are grounded in time-tested concepts from physics, such as symmetries and thermodynamics. PFGM++ takes a century-old idea from fundamental physics — that there might be extra dimensions of space-time — and turns it into a powerful and robust tool to generate synthetic but realistic datasets. I’m thrilled to see the myriad of ways ‘physics intelligence’ is transforming the field of artificial intelligence.”

    The underlying mechanism of PFGM isn’t as complex as it might sound. The researchers compared the data points to tiny electric charges placed on a flat plane in a dimensionally expanded world. These charges produce an “electric field,” with the charges looking to move upwards along the field lines into an extra dimension and consequently forming a uniform distribution on a vast imaginary hemisphere. The generation process is like rewinding a videotape: starting with a uniformly distributed set of charges on the hemisphere and tracking their journey back to the flat plane along the electric lines, they align to match the original data distribution. This intriguing process allows the neural model to learn the electric field, and generate new data that mirrors the original. 

    The PFGM++ model extends the electric field in PFGM to an intricate, higher-dimensional framework. When you keep expanding these dimensions, something unexpected happens — the model starts resembling another important class of models, the diffusion models. This work is all about finding the right balance. The PFGM and diffusion models sit at opposite ends of a spectrum: one is robust but complex to handle, the other simpler but less sturdy. The PFGM++ model offers a sweet spot, striking a balance between robustness and ease of use. This innovation paves the way for more efficient image and pattern generation, marking a significant step forward in technology. Along with adjustable dimensions, the researchers proposed a new training method that enables more efficient learning of the electric field. 

    To bring this theory to life, the team resolved a pair of differential equations detailing these charges’ motion within the electric field. They evaluated the performance using the Frechet Inception Distance (FID) score, a widely accepted metric that assesses the quality of images generated by the model in comparison to the real ones. PFGM++ further showcases a higher resistance to errors and robustness toward the step size in the differential equations.

    Looking ahead, they aim to refine certain aspects of the model, particularly in systematic ways to identify the “sweet spot” value of D tailored for specific data, architectures, and tasks by analyzing the behavior of estimation errors of neural networks. They also plan to apply the PFGM++ to the modern large-scale text-to-image/text-to-video generation.

    “Diffusion models have become a critical driving force behind the revolution in generative AI,” says Yang Song, research scientist at OpenAI. “PFGM++ presents a powerful generalization of diffusion models, allowing users to generate higher-quality images by improving the robustness of image generation against perturbations and learning errors. Furthermore, PFGM++ uncovers a surprising connection between electrostatics and diffusion models, providing new theoretical insights into diffusion model research.”

    “Poisson Flow Generative Models do not only rely on an elegant physics-inspired formulation based on electrostatics, but they also offer state-of-the-art generative modeling performance in practice,” says NVIDIA Senior Research Scientist Karsten Kreis, who was not involved in the work. “They even outperform the popular diffusion models, which currently dominate the literature. This makes them a very powerful generative modeling tool, and I envision their application in diverse areas, ranging from digital content creation to generative drug discovery. More generally, I believe that the exploration of further physics-inspired generative modeling frameworks holds great promise for the future and that Poisson Flow Generative Models are only the beginning.”

    Authors on a paper about this work include three MIT graduate students: Yilun Xu of the Department of Electrical Engineering and Computer Science (EECS) and CSAIL, Ziming Liu of the Department of Physics and the NSF AI IAIFI, and Shangyuan Tong of EECS and CSAIL, as well as Google Senior Research Scientist Yonglong Tian PhD ’23. MIT professors Max Tegmark and Tommi Jaakkola advised the research.

    The team was supported by the MIT-DSTA Singapore collaboration, the MIT-IBM Grand Challenge project, National Science Foundation grants, The Casey and Family Foundation, the Foundational Questions Institute, the Rothberg Family Fund for Cognitive Science, and the ML for Pharmaceutical Discovery and Synthesis Consortium. Their work was presented at the International Conference on Machine Learning this summer. More

  • in

    Meet the 2023-24 Accenture Fellows

    The MIT and Accenture Convergence Initiative for Industry and Technology has selected five new research fellows for 2023-24. Now in its third year, the initiative underscores the ways in which industry and research can collaborate to spur technological innovation.

    Through its partnership with the School of Engineering, Accenture provides five annual fellowships awarded to graduate students with the aim of generating powerful new insights on the convergence of business and technology with the potential to transform society. The 2023-24 fellows will conduct research in areas including artificial intelligence, sustainability, and robotics.

    The 2023-24 Accenture Fellows are:

    Yiyue Luo

    Yiyue Luo is a PhD candidate who is developing innovative integrations of tactile sensing and haptics, interactive sensing and AI, digital fabrication, and smart wearables. Her work takes advantage of recent advances in digital manufacturing and AI, and the convergence in advanced sensing and actuation mechanisms, scalable digital manufacturing, and emerging computational techniques, with the goal of creating novel sensing and actuation devices that revolutionize interactions between people and their environments. In past projects, Luo has developed tactile sensing apparel including socks, gloves, and vests, as well as a workflow for computationally designing and digitally fabricating soft textiles-based pneumatic actuators. With the support of an Accenture Fellowship, she will advance her work of combining sensing and actuating devices and explore the development of haptic devices that simulate tactile cues captured by tactile sensors. Her ultimate aim is to build a scalable, textile-based, closed-loop human-machine interface. Luo’s research holds exciting potential to advance ground-breaking applications for smart textiles, health care, artificial and virtual reality, human-machine interactions, and robotics.

    Zanele Munyikwa is a PhD candidate whose research explores foundation models, a class of models that forms the basis of transformative general-purpose technologies (GPTs) such as GPT4. An Accenture Fellowship will enable Munyikwa to conduct research aimed at illuminating the current and potential impact of foundation models (including large language models) on work and tasks common to “high-skilled” knowledge workers in industries such as marketing, legal services, and medicine, in which foundation models are expected to have significant economic and social impacts. A primary goal of her project is to observe the impact of AI augmentation on tasks like copywriting and long-form writing. A second aim is to explore two primary ways that foundation models are driving the convergence of creative and technological industries, namely: reducing the cost of content generation and enabling the development of tools and platforms for education and training. Munyikwa’s work has important implications for the use of foundation models in many fields, from health care and education to legal services, business, and technology.

    Michelle Vaccaro is a PhD candidate in social engineering systems whose research explores human-AI collaboration with the goals of developing a deeper understanding of AI-based technologies (including ChatGPT and DALL-E), evaluating their performance and evolution, and steering their development toward societally beneficial applications, like climate change mitigation. An Accenture Fellowship will support Vaccaro’s current work toward two key objectives: identifying synergies between humans and AI-based software to help design human-AI systems that address persistent problems better than existing approaches; and investigating applications of human-AI collaboration for forecasting technological change, specifically for renewable energy technologies. By integrating the historically distinct domains of AI, systems engineering, and cognitive science with a wide range of industries, technical fields, and social applications, Vaccaro’s work has the potential to advance individual and collective productivity and creativity in all these areas.

    Chonghuan Wang is a PhD candidate in computational science and engineering whose research employs statistical learning, econometrics theory, and experimental design to create efficient, reliable, and sustainable field experiments in various domains. In his current work, Wang is applying statistical learning techniques such as online learning and bandit theory to test the effectiveness of new treatments, vaccinations, and health care interventions. With the support of an Accenture Fellowship, he will design experiments with the specific aim of understanding the trade-off between the loss of a patient’s welfare and the accuracy of estimating the treatment effect. The results of this research could help to save lives and contain disease outbreaks during pandemics like Covid-19. The benefits of enhanced experiment design and the collection of high-quality data extend well beyond health care; for example, these tools could help businesses optimize user engagement, test pricing impacts, and increase the usage of platforms and services. Wang’s research holds exciting potential to harness statistical learning, econometrics theory, and experimental design in support of strong businesses and the greater social good.

    Aaron Michael West Jr. is a PhD candidate whose research seeks to enhance our knowledge of human motor control and robotics. His work aims to advance rehabilitation technologies and prosthetic devices, as well as improve robot dexterity. His previous work has yielded valuable insights into the human ability to extract information solely from visual displays. Specifically, he demonstrated humans’ ability to estimate stiffness based solely on the visual observation of motion. These insights could advance the development of software applications with the same capability (e.g., using machine learning methods applied to video data) and may enable roboticists to develop enhanced motion control such that a robot’s intention is perceivable by humans. An Accenture Fellowship will enable West to continue this work, as well as new investigations into the functionality of the human hand to aid in the design of a prosthetic hand that better replicates human dexterity. By advancing understandings of human bio- and neuro-mechanics, West’s work has the potential to support major advances in robotics and rehabilitation technologies, with profound impacts on human health and well-being. More

  • in

    How an archeological approach can help leverage biased data in AI to improve medicine

    The classic computer science adage “garbage in, garbage out” lacks nuance when it comes to understanding biased medical data, argue computer science and bioethics professors from MIT, Johns Hopkins University, and the Alan Turing Institute in a new opinion piece published in a recent edition of the New England Journal of Medicine (NEJM). The rising popularity of artificial intelligence has brought increased scrutiny to the matter of biased AI models resulting in algorithmic discrimination, which the White House Office of Science and Technology identified as a key issue in their recent Blueprint for an AI Bill of Rights. 

    When encountering biased data, particularly for AI models used in medical settings, the typical response is to either collect more data from underrepresented groups or generate synthetic data making up for missing parts to ensure that the model performs equally well across an array of patient populations. But the authors argue that this technical approach should be augmented with a sociotechnical perspective that takes both historical and current social factors into account. By doing so, researchers can be more effective in addressing bias in public health. 

    “The three of us had been discussing the ways in which we often treat issues with data from a machine learning perspective as irritations that need to be managed with a technical solution,” recalls co-author Marzyeh Ghassemi, an assistant professor in electrical engineering and computer science and an affiliate of the Abdul Latif Jameel Clinic for Machine Learning in Health (Jameel Clinic), the Computer Science and Artificial Intelligence Laboratory (CSAIL), and Institute of Medical Engineering and Science (IMES). “We had used analogies of data as an artifact that gives a partial view of past practices, or a cracked mirror holding up a reflection. In both cases the information is perhaps not entirely accurate or favorable: Maybe we think that we behave in certain ways as a society — but when you actually look at the data, it tells a different story. We might not like what that story is, but once you unearth an understanding of the past you can move forward and take steps to address poor practices.” 

    Data as artifact 

    In the paper, titled “Considering Biased Data as Informative Artifacts in AI-Assisted Health Care,” Ghassemi, Kadija Ferryman, and Maxine Mackintosh make the case for viewing biased clinical data as “artifacts” in the same way anthropologists or archeologists would view physical objects: pieces of civilization-revealing practices, belief systems, and cultural values — in the case of the paper, specifically those that have led to existing inequities in the health care system. 

    For example, a 2019 study showed that an algorithm widely considered to be an industry standard used health-care expenditures as an indicator of need, leading to the erroneous conclusion that sicker Black patients require the same level of care as healthier white patients. What researchers found was algorithmic discrimination failing to account for unequal access to care.  

    In this instance, rather than viewing biased datasets or lack of data as problems that only require disposal or fixing, Ghassemi and her colleagues recommend the “artifacts” approach as a way to raise awareness around social and historical elements influencing how data are collected and alternative approaches to clinical AI development. 

    “If the goal of your model is deployment in a clinical setting, you should engage a bioethicist or a clinician with appropriate training reasonably early on in problem formulation,” says Ghassemi. “As computer scientists, we often don’t have a complete picture of the different social and historical factors that have gone into creating data that we’ll be using. We need expertise in discerning when models generalized from existing data may not work well for specific subgroups.” 

    When more data can actually harm performance 

    The authors acknowledge that one of the more challenging aspects of implementing an artifact-based approach is being able to assess whether data have been racially corrected: i.e., using white, male bodies as the conventional standard that other bodies are measured against. The opinion piece cites an example from the Chronic Kidney Disease Collaboration in 2021, which developed a new equation to measure kidney function because the old equation had previously been “corrected” under the blanket assumption that Black people have higher muscle mass. Ghassemi says that researchers should be prepared to investigate race-based correction as part of the research process. 

    In another recent paper accepted to this year’s International Conference on Machine Learning co-authored by Ghassemi’s PhD student Vinith Suriyakumar and University of California at San Diego Assistant Professor Berk Ustun, the researchers found that assuming the inclusion of personalized attributes like self-reported race improve the performance of ML models can actually lead to worse risk scores, models, and metrics for minority and minoritized populations.  

    “There’s no single right solution for whether or not to include self-reported race in a clinical risk score. Self-reported race is a social construct that is both a proxy for other information, and deeply proxied itself in other medical data. The solution needs to fit the evidence,” explains Ghassemi. 

    How to move forward 

    This is not to say that biased datasets should be enshrined, or biased algorithms don’t require fixing — quality training data is still key to developing safe, high-performance clinical AI models, and the NEJM piece highlights the role of the National Institutes of Health (NIH) in driving ethical practices.  

    “Generating high-quality, ethically sourced datasets is crucial for enabling the use of next-generation AI technologies that transform how we do research,” NIH acting director Lawrence Tabak stated in a press release when the NIH announced its $130 million Bridge2AI Program last year. Ghassemi agrees, pointing out that the NIH has “prioritized data collection in ethical ways that cover information we have not previously emphasized the value of in human health — such as environmental factors and social determinants. I’m very excited about their prioritization of, and strong investments towards, achieving meaningful health outcomes.” 

    Elaine Nsoesie, an associate professor at the Boston University of Public Health, believes there are many potential benefits to treating biased datasets as artifacts rather than garbage, starting with the focus on context. “Biases present in a dataset collected for lung cancer patients in a hospital in Uganda might be different from a dataset collected in the U.S. for the same patient population,” she explains. “In considering local context, we can train algorithms to better serve specific populations.” Nsoesie says that understanding the historical and contemporary factors shaping a dataset can make it easier to identify discriminatory practices that might be coded in algorithms or systems in ways that are not immediately obvious. She also notes that an artifact-based approach could lead to the development of new policies and structures ensuring that the root causes of bias in a particular dataset are eliminated. 

    “People often tell me that they are very afraid of AI, especially in health. They’ll say, ‘I’m really scared of an AI misdiagnosing me,’ or ‘I’m concerned it will treat me poorly,’” Ghassemi says. “I tell them, you shouldn’t be scared of some hypothetical AI in health tomorrow, you should be scared of what health is right now. If we take a narrow technical view of the data we extract from systems, we could naively replicate poor practices. That’s not the only option — realizing there is a problem is our first step towards a larger opportunity.”  More

  • in

    Helping computer vision and language models understand what they see

    Powerful machine-learning algorithms known as vision and language models, which learn to match text with images, have shown remarkable results when asked to generate captions or summarize videos.

    While these models excel at identifying objects, they often struggle to understand concepts, like object attributes or the arrangement of items in a scene. For instance, a vision and language model might recognize the cup and table in an image, but fail to grasp that the cup is sitting on the table.

    Researchers from MIT, the MIT-IBM Watson AI Lab, and elsewhere have demonstrated a new technique that utilizes computer-generated data to help vision and language models overcome this shortcoming.

    The researchers created a synthetic dataset of images that depict a wide range of scenarios, object arrangements, and human actions, coupled with detailed text descriptions. They used this annotated dataset to “fix” vision and language models so they can learn concepts more effectively. Their technique ensures these models can still make accurate predictions when they see real images.

    When they tested models on concept understanding, the researchers found that their technique boosted accuracy by up to 10 percent. This could improve systems that automatically caption videos or enhance models that provide natural language answers to questions about images, with applications in fields like e-commerce or health care.

    “With this work, we are going beyond nouns in the sense that we are going beyond just the names of objects to more of the semantic concept of an object and everything around it. Our idea was that, when a machine-learning model sees objects in many different arrangements, it will have a better idea of how arrangement matters in a scene,” says Khaled Shehada, a graduate student in the Department of Electrical Engineering and Computer Science and co-author of a paper on this technique.

    Shehada wrote the paper with lead author Paola Cascante-Bonilla, a computer science graduate student at Rice University; Aude Oliva, director of strategic industry engagement at the MIT Schwarzman College of Computing, MIT director of the MIT-IBM Watson AI Lab, and a senior research scientist in the Computer Science and Artificial Intelligence Laboratory (CSAIL); senior author Leonid Karlinsky, a research staff member in the MIT-IBM Watson AI Lab; and others at MIT, the MIT-IBM Watson AI Lab, Georgia Tech, Rice University, École des Ponts, Weizmann Institute of Science, and IBM Research. The paper will be presented at the International Conference on Computer Vision.

    Focusing on objects

    Vision and language models typically learn to identify objects in a scene, and can end up ignoring object attributes, such as color and size, or positional relationships, such as which object is on top of another object.

    This is due to the method with which these models are often trained, known as contrastive learning. This training method involves forcing a model to predict the correspondence between images and text. When comparing natural images, the objects in each scene tend to cause the most striking differences. (Perhaps one image shows a horse in a field while the second shows a sailboat on the water.)

    “Every image could be uniquely defined by the objects in the image. So, when you do contrastive learning, just focusing on the nouns and objects would solve the problem. Why would the model do anything differently?” says Karlinsky.

    The researchers sought to mitigate this problem by using synthetic data to fine-tune a vision and language model. The fine-tuning process involves tweaking a model that has already been trained to improve its performance on a specific task.

    They used a computer to automatically create synthetic videos with diverse 3D environments and objects, such as furniture and luggage, and added human avatars that interacted with the objects.

    Using individual frames of these videos, they generated nearly 800,000 photorealistic images, and then paired each with a detailed caption. The researchers developed a methodology for annotating every aspect of the image to capture object attributes, positional relationships, and human-object interactions clearly and consistently in dense captions.

    Because the researchers created the images, they could control the appearance and position of objects, as well as the gender, clothing, poses, and actions of the human avatars.

    “Synthetic data allows a lot of diversity. With real images, you might not have a lot of elephants in a room, but with synthetic data, you could actually have a pink elephant in a room with a human, if you want,” Cascante-Bonilla says.

    Synthetic data have other advantages, too. They are cheaper to generate than real data, yet the images are highly photorealistic. They also preserve privacy because no real humans are shown in the images. And, because data are produced automatically by a computer, they can be generated quickly in massive quantities.

    By using different camera viewpoints, or slightly changing the positions or attributes of objects, the researchers created a dataset with a far wider variety of scenarios than one would find in a natural dataset.

    Fine-tune, but don’t forget

    However, when one fine-tunes a model with synthetic data, there is a risk that model might “forget” what it learned when it was originally trained with real data.

    The researchers employed a few techniques to prevent this problem, such as adjusting the synthetic data so colors, lighting, and shadows more closely match those found in natural images. They also made adjustments to the model’s inner-workings after fine-tuning to further reduce any forgetfulness.

    Their synthetic dataset and fine-tuning strategy improved the ability of popular vision and language models to accurately recognize concepts by up to 10 percent. At the same time, the models did not forget what they had already learned.

    Now that they have shown how synthetic data can be used to solve this problem, the researchers want to identify ways to improve the visual quality and diversity of these data, as well as the underlying physics that makes synthetic scenes look realistic. In addition, they plan to test the limits of scalability, and investigate whether model improvement starts to plateau with larger and more diverse synthetic datasets.

    This research is funded, in part, by the U.S. Defense Advanced Research Projects Agency, the National Science Foundation, and the MIT-IBM Watson AI Lab. More

  • in

    Fast-tracking fusion energy’s arrival with AI and accessibility

    As the impacts of climate change continue to grow, so does interest in fusion’s potential as a clean energy source. While fusion reactions have been studied in laboratories since the 1930s, there are still many critical questions scientists must answer to make fusion power a reality, and time is of the essence. As part of their strategy to accelerate fusion energy’s arrival and reach carbon neutrality by 2050, the U.S. Department of Energy (DoE) has announced new funding for a project led by researchers at MIT’s Plasma Science and Fusion Center (PSFC) and four collaborating institutions.

    Cristina Rea, a research scientist and group leader at the PSFC, will serve as the primary investigator for the newly funded three-year collaboration to pilot the integration of fusion data into a system that can be read by AI-powered tools. The PSFC, together with scientists from William & Mary, the University of Wisconsin at Madison, Auburn University, and the nonprofit HDF Group, plan to create a holistic fusion data platform, the elements of which could offer unprecedented access for researchers, especially underrepresented students. The project aims to encourage diverse participation in fusion and data science, both in academia and the workforce, through outreach programs led by the group’s co-investigators, of whom four out of five are women. 

    The DoE’s award, part of a $29 million funding package for seven projects across 19 institutions, will support the group’s efforts to distribute data produced by fusion devices like the PSFC’s Alcator C-Mod, a donut-shaped “tokamak” that utilized powerful magnets to control and confine fusion reactions. Alcator C-Mod operated from 1991 to 2016 and its data are still being studied, thanks in part to the PSFC’s commitment to the free exchange of knowledge.

    Currently, there are nearly 50 public experimental magnetic confinement-type fusion devices; however, both historical and current data from these devices can be difficult to access. Some fusion databases require signing user agreements, and not all data are catalogued and organized the same way. Moreover, it can be difficult to leverage machine learning, a class of AI tools, for data analysis and to enable scientific discovery without time-consuming data reorganization. The result is fewer scientists working on fusion, greater barriers to discovery, and a bottleneck in harnessing AI to accelerate progress.

    The project’s proposed data platform addresses technical barriers by being FAIR — Findable, Interoperable, Accessible, Reusable — and by adhering to UNESCO’s Open Science (OS) recommendations to improve the transparency and inclusivity of science; all of the researchers’ deliverables will adhere to FAIR and OS principles, as required by the DoE. The platform’s databases will be built using MDSplusML, an upgraded version of the MDSplus open-source software developed by PSFC researchers in the 1980s to catalogue the results of Alcator C-Mod’s experiments. Today, nearly 40 fusion research institutes use MDSplus to store and provide external access to their fusion data. The release of MDSplusML aims to continue that legacy of open collaboration.

    The researchers intend to address barriers to participation for women and disadvantaged groups not only by improving general access to fusion data, but also through a subsidized summer school that will focus on topics at the intersection of fusion and machine learning, which will be held at William & Mary for the next three years.

    Of the importance of their research, Rea says, “This project is about responding to the fusion community’s needs and setting ourselves up for success. Scientific advancements in fusion are enabled via multidisciplinary collaboration and cross-pollination, so accessibility is absolutely essential. I think we all understand now that diverse communities have more diverse ideas, and they allow faster problem-solving.”

    The collaboration’s work also aligns with vital areas of research identified in the International Atomic Energy Agency’s “AI for Fusion” Coordinated Research Project (CRP). Rea was selected as the technical coordinator for the IAEA’s CRP emphasizing community engagement and knowledge access to accelerate fusion research and development. In a letter of support written for the group’s proposed project, the IAEA stated that, “the work [the researchers] will carry out […] will be beneficial not only to our CRP but also to the international fusion community in large.”

    PSFC Director and Hitachi America Professor of Engineering Dennis Whyte adds, “I am thrilled to see PSFC and our collaborators be at the forefront of applying new AI tools while simultaneously encouraging and enabling extraction of critical data from our experiments.”

    “Having the opportunity to lead such an important project is extremely meaningful, and I feel a responsibility to show that women are leaders in STEM,” says Rea. “We have an incredible team, strongly motivated to improve our fusion ecosystem and to contribute to making fusion energy a reality.” More