In the three years since ChatGPT's explosive debut, OpenAI's technology has upended a remarkable range of everyday activities at home, at work, and in schools. Now OpenAI is making an explicit play for scientists. In October, the firm announced that it had launched a whole new team, called OpenAI for Science, dedicated to exploring how its large language models could help scientists and tweaking its tools to support them. How does a push into science fit with OpenAI's wider mission? And what exactly is the firm hoping to achieve? I put these questions to Kevin Weil, a vice president at OpenAI who leads the new OpenAI for Science team, in an exclusive interview.

Russia has been exploiting Ukraine's harshest winter in years to pummel energy infrastructure across the country. Repeated strikes have crippled the power supply to major Ukrainian cities, leaving millions without heating or light as temperatures hover around -15C (5F) for the third week in a row. Electrical companies carry out round-the-clock repairs - only for their work to be undone at night, when Russian drone and missiles again damage power stations. In Kyiv, people were initially able to keep the cold at bay by using electric heaters or wrapping up warm. But the freezing temperatures have lasted weeks now, with no end in sight.

Shaping multi-megawatt loads, such as data centers, impacts generator dispatch on the electric grid, which in turn affects system CO2 emissions and energy cost. Substantiating the effectiveness of prevalent load shaping strategies, such as those based on grid-level average carbon intensity, locational marginal price, or marginal emissions, is challenging due to the lack of detailed counterfactual data required for accurate attribution. This study uses a series of calibrated granular ERCOT day-ahead direct current optimal power flow (DC-OPF) simulations for counterfactual analysis of a broad set of load shaping strategies on grid CO2 emissions and cost of electricity. In terms of annual grid level CO2 emissions reductions, LMP-based shaping outperforms other common strategies, but can be significantly improved upon. Examining the performance of practicable strategies under different grid conditions motivates a more effective load shaping approach: one that "cherry-picks" a daily strategy based on observable grid signals and historical data. The cherry-picking approach to power load shaping is applicable to any large flexible consumer on the electricity grid, such as data centers, distributed energy resources and Virtual Power Plants (VPPs).

Load balancing-the allocation of work across parallel resources to reduce delay, energy and cost-is a pervasive challenge in science and engineering, from large-scale simulation and data processing to cloud and manufacturing operations. Motivated by the emerging bottleneck in large language model (LLM) serving, we study a particularly stringent regime of load balancing that arises in barrier-synchronized, stateful systems: work cannot be freely migrated and progress is gated by the slowest participant at each step, so heterogeneity and temporal drift in workloads create persistent stragglers and substantial idle time. LLM serving under data-parallel decoding provides a prominent modern instance: in production traces, barrier-induced idle can exceed 40% of compute time per decode step. Here we develop a universal load-balancing principle, which admits a step-wise finite-horizon integer-optimization formulation and yields worst-case guarantees: across LLM decode models and a broader class of non-decreasing workload drift processes, it reduces long-run imbalance by a factor that grows with batch size and system scale. Extensive experiments corroborate the theory, showing substantial improvements in throughput and latency together with reductions in energy consumption. These results provide a general, theoretically grounded framework for load balancing, with immediate implications for sustainable LLM serving and broad relevance to other synchronization-gated resource-allocation problems.

Environment Animals Wildlife Fish'Walking sharks' lay eggs without breaking a sweat Breakthroughs, discoveries, and DIY tips sent six days a week. Being pregnant and giving birth is hard work for any species--but epaulette sharks () might disagree. These fish and a number of other species are known as " walking sharks " for their ability to traverse both the seafloor and land with their fins. Epaulette sharks' energy use didn't change during their reproduction cycle, as described in a study recently published in the journal . "Reproduction is the ultimate investment you are literally building new life from scratch," Jodie Rummer, a marine biologist at James Cook University and co-author of the recent study, said in a university statement .

This material may not be published, broadcast, rewritten, or redistributed. Quotes displayed in real-time or delayed by at least 15 minutes. Market data provided by Factset . Powered and implemented by FactSet Digital Solutions . Mutual Fund and ETF data provided by LSEG . 'Are You Dead?' app taps into global loneliness crisis Can autonomous trucks really make highways safer?

A cordless, portable blender was barely possible a few years ago. But two cordless blenders are ahead of the pack. But battery tech keeps getting better. This means the best portable blender I've tested, the Ninja Blast Max ($100), is now fully able to make a six-pack of crushed-ice margaritas at your next picnic or blend up a berry-filled protein shake at the gym without breaking much of a sweat. Meanwhile, the ingeniously designed Nutribullet Flip ($115) offers more torque than previous-generation blenders, plus enough insulation to keep ice frozen until it's time for lunch (or even dinner).

Shipping experts say the big winter storm across a wide swath of the country should be business as usual--if their safeguards hold. Up to two-thirds of the US is facing down the threat of serious snow, cold, and ice this weekend, with the potential to snarl roads (and the businesses that depend on them) from Texas up to New York City . At this point, grocery stores, logistics experts, warehouse operators, and trucking companies have been prepping for days. Still, the effects on the supply chain--and the retail store shelves that depend on them--are yet to be determined. On one hand, this is winter business as usual.

Could Ukraine hold a presidential election right now? Will Europe use frozen Russian assets to fund war? How can Ukraine rebuild China ties? 'Ukraine is running out of men, money and time' Russia targets Ukraine's energy as trilateral talks loom As the presidents of Ukraine, Russia and the United States prepare to hold their first trilateral meeting to end Russia's war in Ukraine this weekend, almost half of Ukraine is without electricity and heat in sub-zero temperatures, following repeated Russian drone strikes targeting energy infrastructure. The strikes appeared designed to break Ukrainian resistance at the negotiating table on territorial concessions to Russia - the one issue Ukraine and the US said remained unresolved at the end of talks in Davos, Switzerland, between Ukraine's Volodymyr Zelenskyy and US President Donald Trump this week.

Bayesian optimisation is a sample efficient method for finding a global optimum of expensive black-box objective functions. Historic datasets from related problems can be exploited to help improve performance of Bayesian optimisation by adapting transfer learning methods to various components of the Bayesian optimisation pipeline. In this study we perform an empirical analysis of various ensemble-based transfer learning Bayesian optimisation methods and pipeline components. We expand on previous work in the literature by contributing some specific pipeline components, and three new real-time transfer learning Bayesian optimisation benchmarks. In particular we propose to use a weighting strategy for ensemble surrogate model predictions based on regularised regression with weights constrained to be positive, and a related component for handling the case when transfer learning is not improving Bayesian optimisation performance. We find that in general, two components that help improve transfer learning Bayesian optimisation performance are warm start initialisation and constraining weights used with ensemble surrogate model to be positive.