Marijn Heule argues LLMs dissolve the encoding bottleneck in computer-aided mathematics, making formal proofs accessible to far more researchers. A new CACM piece and fresh education studies urge the field to demand released models, full failure reporting, and harder real-world problems. Transparency will determine if this lowers barriers or truly expands discovery.
| # | Наименование новости | Тональность | Информативность | Дата публикации |
|---|---|---|---|---|
| 1 | Les mathématiciens se positionnent face à l’IA | 0 | 14.79 | 09-06-2026 |
| 2 | Hidden goals can undermine AI teamwork, study finds | 0 | 8.24 | 06-08-2026 |
| 3 | Как заставить LLM считать точно: генерация кода вместо генерации ответов | 0 | 11.55 | 28-03-2026 |
| 4 | Les mathématiques dopées à l’IA | 0 | 8.77 | 05-03-2026 |
| 5 | Generative AI has changed mathematics forever. Where to from here? | 0 | 12.18 | 10-08-2026 |
| 6 | A new 'golden age' of mathematics may be dawning, thanks to AI and human ingenuity | 0 | 9.12 | 23-07-2026 |
| 7 | OpenAI says next-generation model solved 10 longstanding math problems | 0 | 12.42 | 03-08-2026 |
| 8 | Я устал писать одноразовые скрипты для бенчмарков LLM и собрал харнесс, который сам считает Pareto-front | 0 | 7 | 27-06-2026 |
| 9 | College startup takes aim at research fraud | 0 | 18.4 | 12-08-2026 |