
GPT-6 Astra completed a 50-tab tax workbook twice as fast as GPT-5.6 Sol, and its stronger understanding of user intent gives Basis more confidence in real-world use.
A startup that builds AI agents for accounting work tested a newer AI model on complex tax tasks. The newer model completed a 50-tab tax workbook in half the time of the previous version and showed better understanding of what users were trying to accomplish. The model can also adjust how much computational effort it uses depending on task difficulty, which reduces costs and response time for long tasks. Internal evaluation scores improved by about 20 percent, suggesting the model can handle situations beyond those it was explicitly trained on.

Deep Blue took down Garry Kasparov at chess in 1997, AlphaGo beat Lee Sedol at Go in 2016, and poker bots have been beating professionals for years. But one classic game called Stratego held out. Even DeepMind, with its exceptional budget, couldn't build a machine that reliably beat the best human players. Now, a team of researchers from Carnegie Mellon, MIT, New York University, and Stanford University has done it. Their AI, called Ataraxos, beat Pim Niemeijer, arguably the best Stratego player

Opus 5.5’s biggest tell is the word “dependable,” which pops up 23 times more often than in human samples.
Comparing speech synthesis systems across languages and voices has lacked standardized metrics, making it harder to track progress in this rapidly advancing field.
Want to go deeper than the news? Explore live, cohort-based AI courses taught by practitioners.
Browse AI courses on Maven