Google launched Gemini Decision 1 on 11 November with a benchmark table it made itself. The tableβs own cost row puts TypeSafeβs Jev below it, the βnearly 5x cheaperβ line is measured against another Gemini model, and Artificial Analysisβs run the same day lands it mid-pack, at more than twice Jevβs cost per task. The latency claim held up.
FABench backtests forecasting agents at 1 July 2026, and the leakage controls are the best part of it. The headline, that the leading agents match the prediction market, rests on 54 questions. Most of the matched contracts settle under different rules, and the markets have a median lifetime volume of $10.4 thousand. On the full set, the best agent removes about 31 percent of a coin flipβs squared error.
Mask a language model down to five answer letters and the softmax hands back numbers like 0.9988. That is a probability conditional on the mask, and the mask throws away a quantity that bears on how far to trust it.
Microsoft-Decision-1 and H2O-Lightning-4B both launched against JevBench within three days. One is billed as #1 on the board, the other as the fastest model measured. Neither claim is quite what the leaderboard says.
A nine-author paper models LLM adoption as a contagion and finds tipping points, lock-in, and βabrupt losses in cognitive competence.β The compartment structure needs five rate parameters. The figures assign four of them fixed values with no fitting procedure and no dataset, because no dataset exists. The tipping point is a property of the functional form, and the paper says so β the risk is in who cites it next.
The statistic that launched a thousand data platforms β β80% of enterprise data is unstructuredβ β turns out to be a hedged guess from a 1998 analyst note. The real story is better: AI can now read documents at human level, still struggles to extract from them faithfully, and the honest evidence on what that unlocks for decisions.
A digital twin is not a fancy 3D model β it is a live data link between a virtual model and one specific physical thing. By the literatureβs own classification, most products sold as twins arenβt. What the research actually says, which famous numbers are marketing, and why synchronization is the whole game.
On classroom benchmarks, AI writes SQL better than most analysts. On real enterprise warehouses, the best published systems still fail most tasks end-to-end. The gap between those two sentences is the most important number in analytics β and the evidence about what closes it.
Decision intelligence is the discipline of engineering the decision, not just the data β and its scientific foundations are seventy years old, awkward, and mostly ignored. What the evidence actually says about when human-plus-AI beats either alone, and why the answer is βless often than youβd hope, unless you design for it.β
Before any lander is cleared to hit an atmosphere, engineers fly the mission thousands of times in software, each run with a different set of imperfections. How dispersion analysis, 3-sigma design, and importance sampling turned landing on Mars from a gamble into an engineering discipline.