The Download: Trump's impact on science, and meet our climate and energy honorees How Trump's policies are affecting early-career scientists--in their own words Every year MIT Technology Review celebrates accomplished young scientists, entrepreneurs, and inventors from around the world in our Innovators Under 35 list. We've just published the 2025 edition . This year, though, the context is different: The US scientific community is under attack. Since Donald Trump took office in January, his administration has fired top government scientists, targeted universities and academia, and made substantial funding cuts to the country's science and technology infrastructure. We asked our six most recent cohorts about both positive and negative impacts of the administration's new policies. Their responses provide a glimpse into the complexities of building labs, companies, and careers in today's political climate.

After growing up without reliable power at home, Iwnetim Abate is working to develop a steady supply of sustainable energy. "I'm the only one who wears glasses and has eye problems in the family," Iwnetim Abate says with a smile as sun streams in through the windows of his MIT office. "I think it's because of the candles." In the small town in Ethiopia where he grew up, Abate's family had electricity, but it was unreliable. So, for several days each week when they were without power, Abate would finish his homework by candlelight. Today, Abate, 32, is an assistant professor at MIT in the department of materials science and engineering.

In this paper we propose a formal, model-agnostic meta-learning framework for safe reinforcement learning. Our framework is inspired by how parents safeguard their children across a progression of increasingly riskier tasks, imparting a sense of safety that is carried over from task to task. We model this as a meta-learning process where each task is synchronized with a safeguard that monitors safety and provides a reward signal to the agent. The safeguard is implemented as a finite-state machine based on a safety specification; the reward signal is formally shaped around this specification. The safety specification and its corresponding safeguard can be arbitrarily complex and non-Markovian, which adds flexibility to the training process and explainability to the learned policy. The design of the safeguard is manual but it is high-level and model-agnostic, which gives rise to an end-to-end safe learning approach with wide applicability, from pixel-level game control to language model fine-tuning. Starting from a given set of safety specifications (tasks), we train a model such that it can adapt to new specifications using only a small number of training samples. This is made possible by our method for efficiently transferring safety bias between tasks, which effectively minimizes the number of safety violations. We evaluate our framework in a Minecraft-inspired Gridworld, a VizDoom game environment, and an LLM fine-tuning application. Agents trained with our approach achieve near-minimal safety violations, while baselines are shown to underperform.

This paper addresses the problem of maintaining safety during training in Reinforcement Learning (RL), such that the safety constraint violations are bounded at any point during learning. In a variety of RL applications the safety of the agent is particularly important, e.g. autonomous platforms or robots that work in proximity of humans. As enforcing safety during training might severely limit the agent's exploration, we propose here a new architecture that handles the trade-off between efficient progress and safety during exploration. As the exploration progresses, we update via Bayesian inference Dirichlet-Categorical models of the transition probabilities of the Markov decision process that describes the environment dynamics. This paper proposes a way to approximate moments of belief about the risk associated to the action selection policy. We construct those approximations, and prove the convergence results. We propose a novel method for leveraging the expectation approximations to derive an approximate bound on the confidence that the risk is below a certain level. This approach can be easily interleaved with RL and we present experimental results to showcase the performance of the overall architecture.

Controllers for dynamical systems that operate in safety-critical settings must account for stochastic disturbances. Such disturbances are often modeled as process noise in a dynamical system, and common assumptions are that the underlying distributions are known and/or Gaussian. In practice, however, these assumptions may be unrealistic and can lead to poor approximations of the true noise distribution. We present a novel controller synthesis method that does not rely on any explicit representation of the noise distributions. In particular, we address the problem of computing a controller that provides probabilistic guarantees on safely reaching a target, while also avoiding unsafe regions of the state space. First, we abstract the continuous control system into a finite-state model that captures noise by probabilistic transitions between discrete states. As a key contribution, we adapt tools from the scenario approach to compute probably approximately correct (PAC) bounds on these transition probabilities, based on a finite number of samples of the noise. We capture these bounds in the transition probability intervals of a so-called interval Markov decision process (iMDP). This iMDP is, with a user-specified confidence probability, robust against uncertainty in the transition probabilities, and the tightness of the probability intervals can be controlled through the number of samples. We use state-of-the-art verification techniques to provide guarantees on the iMDP and compute a controller for which these guarantees carry over to the original control system. In addition, we develop a tailored computational scheme that reduces the complexity of the synthesis of these guarantees on the iMDP. Benchmarks on realistic control systems show the practical applicability of our method, even when the iMDP has hundreds of millions of transitions.

Controllers for dynamical systems that operate in safety-critical settings must account for stochastic disturbances. Such disturbances are often modeled as process noise in a dynamical system, and common assumptions are that the underlying distributions are known and/or Gaussian. In practice, however, these assumptions may be unrealistic and can lead to poor approximations of the true noise distribution. We present a novel controller synthesis method that does not rely on any explicit representation of the noise distributions. In particular, we address the problem of computing a controller that provides probabilistic guarantees on safely reaching a target, while also avoiding unsafe regions of the state space. First, we abstract the continuous control system into a finite-state model that captures noise by probabilistic transitions between discrete states. As a key contribution, we adapt tools from the scenario approach to compute probably approximately correct (PAC) bounds on these transition probabilities, based on a finite number of samples of the noise. We capture these bounds in the transition probability intervals of a so-called interval Markov decision process (iMDP). This iMDP is, with a user-specified confidence probability, robust against uncertainty in the transition probabilities, and the tightness of the probability intervals can be controlled through the number of samples. We use state-of-the-art verification techniques to provide guarantees on the iMDP and compute a controller for which these guarantees carry over to the original control system. In addition, we develop a tailored computational scheme that reduces the complexity of the synthesis of these guarantees on the iMDP. Benchmarks on realistic control systems show the practical applicability of our method, even when the iMDP has hundreds of millions of transitions.

Dr. Sihao Solar discusses his award-winning analysis within the space of controlling the flight of a drone when confronted with a sudden rotor failure. Typical analysis on this space addressed the case the place one of many 4 rotors in a quadrotor all of a sudden, spontaneously stops working. This earlier analysis doesn't take into full account real-life eventualities the place rotor failure is frequent. This consists of collisions with different drones, partitions, birds, and working in degraded GPS environments. Dr. Sihao Solar is a postdoctoral analysis assistant on the Robotics and Notion Group (RPG) in College of Zurich directed by Prof. Davide Scaramuzza.

Machine Learning at the edge is gaining steam. BrainChip is accelerating this with their Akida architecture, which is mimicking the human brain by incorporating the 5 human senses on a machine learning-enabled chip. Their chips will let roboticists and IoT developers run ML on device for low latency, low power, and low-cost machine learning-enabled products. This opens up a new product category where everyday devices can affordably become smart devices. Rob is an AI thought-leader and Vice President of Worldwide Sales at BrainChip, a global tech company that has developed artificial intelligence that learns like a brain, whilst prioritizing efficiency, ultra-low power consumption, and continuous learning. Rob has over 20 years of sales expertise in licensing intellectual property and selling EDA technology and attended Harvard Business School. This is your host Abate, founder of fluid dev a platform that helps robotics and machine learning companies scale their teams up as they grow. So welcome Rob and honor to have you on here. Rob: Abate it's great to be here and thank you for having me on your podcast.

Truly "autonomous" systems are starting to replace or augment many of the routine tasks and processes people perform every day, improving efficiency while freeing individuals for higher-level pursuits. But what's often overlooked is how much progress is happening in other areas and industries: healthcare, air travel, energy provision, retail, logistics, agriculture, and construction. Autonomous systems are even helping governments match refugees with the most suitable communities to live, as detailed in one of the four real-world vignettes we present below. Such optimism makes sense, given advances such as self-managing and self-patching databases in IT. But our survey's other findings might underestimate the pace of change: Just 24% say they expect to see significant use of autonomous tech in construction, for example, even though self-driving bulldozers already are in use on select projects.

This paper proposes the first model-free Reinforcement Learning (RL) framework to synthesise policies for an unknown, and possibly continuous-state, Markov Decision Process (MDP), such that a given linear temporal property is satisfied. We convert the given property into a Limit Deterministic Buchi Automaton (LDBA), namely a finite-state machine expressing the property. Exploiting the structure of the LDBA, we shape an adaptive reward function on-the-fly, so that an RL algorithm can synthesise a policy resulting in traces that probabilistically satisfy the linear temporal property. This probability (certificate) is also calculated in parallel with learning, i.e. the RL algorithm produces a policy that is certifiably safe with respect to the property. Under the assumption that the MDP has a finite number of states, theoretical guarantees are provided on the convergence of the RL algorithm. We also show that our method produces "best available" control policies when the logical property cannot be satisfied. Whenever the MDP has a continuous state space, we empirically show that our framework finds satisfying policies, if there exist such policies. Additionally, the proposed algorithm can handle time-varying periodic environments. The performance of the proposed architecture is evaluated via a set of numerical examples and benchmarks, where we observe an improvement of one order of magnitude in the number of iterations required for the policy synthesis, compared to existing approaches whenever available.