A blog post examines C17 section 7.25.2.6, which governs error handling for strtod, strtof, strtold and related string-to-float conversion functions. It explains that these functions' ability to report overflow or underflow via errno depends on the math_errhandling macro, and that on libraries like musl where MATH_ERRNO is not advertised, there is no standard-compliant way to detect these conversion errors. The author notes that Linux man pages for these functions omit this detail entirely.
sebsite.pw
· 2026-10-10
OpenAI published more than 700 preliminary math research papers on GitHub this week, intensifying tension with academic mathematicians. The release follows an open letter last month from 28 Fields Medal winners who created a site called Math and AI, accusing AI companies of pursuing goals misaligned with the mathematics community.
cnet.com
· 2026-10-09
In a guest post on Terence Tao's blog, mathematics professor Álvaro Lozano-Robledo responds to a distressed undergraduate who questioned whether a career in math academia is still viable given rapid progress in large language models. The student, who described feeling at a loss about career trajectory and even the meaning of life, asked whether academia would remain open to passionate but not exceptionally gifted students. Lozano-Robledo frames his post as advice to reassure students making near-term decisions about their futures.
terrytao.wordpress.com
· 2026-10-09
OpenAI revised its public GitHub repository of formalized mathematical results, adding six new Lean formalizations and modifying 19 existing ones. It also withdrew three results previously listed as formalized. The company said roughly 42% of its top-line results are now formalized and that it will keep updating the repo as it finds errors.
twitter.com
· 2026-10-08
OpenAI has published a large set of new results on mathematical problem-solving benchmarks, adding to the growing body of data on AI capabilities in this area. The release includes hundreds of new test outcomes, highlighting progress and ongoing challenges in AI understanding of complex math tasks.
science.slashdot.org
· 2026-10-08
OpenAI has published results claiming its AI systems made progress on more than 300 mathematical problems, following an earlier claimed solution related to a Millennium Prize problem. The company presented the findings as an attempt to demonstrate credibility to the mathematics research community.
wsj.com
· 2026-10-07
Cleo, a user on Math.SE, successfully derived a closed-form solution for a challenging integral that had stumped others, expressing it in terms of the golden ratio. Her answer was posted shortly after another user’s detailed contour integral solution, but lacked supporting proof. The community questioned the answer’s completeness due to the absence of detailed derivation.
en.wikipedia.org
· 2026-10-06
Stephen Wolfram published commentary and a video addressing claims that AI systems solving math problems could make human mathematicians obsolete. He draws on his experience introducing Mathematica in 1988, arguing similar predictions arose then but instead elevated the level of math people could do rather than replacing mathematicians. Wolfram contends pure mathematics is not merely about solving problems like symbolic integration but is a centuries-old intellectual tradition dating to Plato and Euclid.
writings.stephenwolfram.com
· 2026-10-04
A technical post explains the Softmax function, which converts an N-dimensional vector of real numbers into another N-dimensional vector of values between 0 and 1 that sum to 1. Using worked numeric examples, it shows how Softmax acts as a smoothed version of the maximum function, giving the largest input a disproportionately larger share of the output while still assigning smaller shares to other elements.
eli.thegreenplace.net
· 2026-10-01
A developer has released Lathoa, a math learning app for children built around AI-generated answers that are intentionally incorrect. Children progress through a ranking system—Rookie, Inspector, Detective, Sherlock—by catching the AI's mistakes rather than by speed or grades, with leaderboards optional rather than default.
lathoa.ai
· 2026-09-30
Modern chess engines can defeat the world's best grandmasters, but researchers note they still cannot converge on one definitive, infallible way to play the game. Rather than making chess feel 'solved' or stale, these AI systems are reportedly revealing new strategic ideas and approaches to the game.
wsj.com
· 2026-09-23
A developer building the habit-tracker app Main Character reported reaching 1,000 consecutive days of daily math practice on Math Academy, progressing from Math Foundations through Mathematics for Machine Learning. They disclosed that relying on notes and answer-checking during difficult lessons masked underlying weaknesses, prompting them to reset their course progress and restart.
gmays.com
· 2026-09-23
GoKawiil