AI companies pivot to science as coding benchmarks get boring
Original headline: Moonshot AI pivots Kimi models from coding to science as benchmark gains flatten — the-decoder.com
More in AI and machine learning news
Chinese startup Moonshot AI is shifting its Kimi language models from writing code to tackling scientific research in biology, chemistry, and pharmaceuticals. International director Giacomo Ficari explained at Web Summit that science requires more sophisticated AI than fields like law, finance, or programming, suggesting the coding benchmarks have become too easy or saturated to show meaningful progress.
Despite one demo where Kimi analyzed thousands of mouse pancreatic cells faster than human researchers, the broader picture shows limited real impact. Major models from Anthropic and OpenAI have faced criticism for overhyped math and physics claims, while a pharmaceutical industry survey reveals deep skepticism—only 24% expect AI to significantly improve drug development success rates. The pivot to science looks less like a breakthrough and more like chasing the next unproven hype cycle after coding benchmarks plateaued.
Read the original at the-decoder.com
Satirical commentary on real tech news, summarised with AI. Read the original source for the facts. How ZeroDaily works
