Topic
Everything filed under AI, newest first.
RSS · JSON · All topics
Amazon saves millions by replacing Sonnet 4.6 with Strands Decider
Two Amazon engineers replaced Claude Sonnet 4.6 with Strands Decider, an open decision model, for ticket labeling in one of the world's largest internal ticketing systems; it performed as well at about 98% less cost, and they explain running both models side by side on real traffic before switching.
4 min · 1,011 words
Update on October 2026 security incident
Fireworks AI updates its October 2026 security bulletin: an LLMJacking-linked threat actor accessed account and billing information and, for some accounts, secrets, datasets and incomplete model weights, partly via a breached third-party cloud provider; Fireworks details remediation with Palo Alto Networks Unit 42.
3 min · 588 words
Why is AI Slop so Frustrating to Read?
Zonder Labs argues most workplace AI slop is unclear writing rather than stylistic tics like em dashes, defines three clarity faults (hidden characters and actions, vague wording, empty statements and misplaced emphasis), and measures them in AI vs human articles and Claude Code vs Stack Overflow answers.
5 min · 1,193 words
Introducing Claude Fable 5.2 and Claude Mythos 5.2
Anthropic upgrades Claude Fable and Claude Mythos, the same underlying model with different safeguards, to version 5.2, citing gains in coding, reasoning and vision, plus prerelease results like designing working electronics and formalizing Fermat's Last Theorem.
26 min · 6,060 words
Testing Claude on real-world electronics design
Anthropic and Diode Computers asked Claude to design three electronic devices, a speakerphone, a voice-assistant hub and a handheld virtual pet, from start to finish. The prototypes worked without any fixes to their circuitry.
13 min · 3,002 words
Our newest TPUs: 8tx and 8i Max, built for the agentic era
Google expands its eighth-generation TPU portfolio with TPU 8tx, aimed at secure inference, and TPU 8i Max, built for massive context windows, as it scales infrastructure for agentic AI workloads.
8 min · 1,829 words
Marcelo Eduardo explains how he trained small open image models with LoRAs to paint isometric pixel-art cities the way Codex painted his Brasília map, cutting the cost of maps like MiniDF and MiniSP from a $500-a-month Codex plan to a rented GPU or a home computer.
3 min · 618 words
What is the superior coffee order? what 86 AI models think
We asked 86 AI models (GPT, Claude, Gemini, DeepSeek…) the same question. 34% picked Flat White. See every answer, who searched the web, and who dissented.
9 min · 2,122 words
Quadrillion-parameter transformers and how to train them
Q Labs researchers argue that closing AI's sample-efficiency gap means training far larger models on fixed data, and show that bigger models on repeated data need much more weight decay; their hyperparameter law stays near-optimal at 16x the fitted compute, and with ensembling loss keeps falling as compute grows.
7 min · 1,668 words
The future of Software Engineering is boring
Alex Pliutau argues that AI coding has turned much of software engineering into caretaking code nobody fully understands, and that the work may get more boring before it gets better, with human-authored code eventually reserved for hobbies or high-criticality systems.
6 min · 1,269 words
A longtime Claude Code user explains why an accumulation of small frustrations, from slow AGENTS.md support and single-session IDE integration to constant system-prompt churn and heavy tool-description overhead, pushed him to retire it as his main coding harness and move his development setup to the open Pi coding agent.
4 min · 1,022 words
On Chomsky and the Two Cultures of Statistical Learning
Peter Norvig responds to Noam Chomsky's 2011 criticism of purely statistical approaches to language, explaining what statistical models are, documenting their success in language tasks, and arguing that probabilistic modelling is a legitimate scientific way to understand language rather than mere mimicry.
37 min · 8,521 words
What objective is reasoning solving?
Alex Wang frames LLM reasoning as pathfinding in token space, comparing best-of-k sampling with training models to generate better samples, and uses that lens to discuss amortization, unfaithful chains of thought, and reward hacking from both verifier and sampler.
10 min · 2,275 words
I'm Not Convinced The AI Bubble Exists
Matteo Raso argues the AI industry may not be a bubble: running costs are constrained, revenues like those in Anthropic's S-1 are growing fast, switching costs act as a moat, capital remains plentiful, and public AI company valuations don't resemble past bubbles, while acknowledging real risks such as debt and data-centre backlash.
12 min · 2,851 words
Wilson Tovar reflects on binge-watching the TV series Person of Interest 15 years after it aired, how it brought back memories of early-2010s internet culture such as SOPA/PIPA and the Megaupload shutdown, and how close its all-seeing Machine now feels to today's surveillance and AI reality.
1 min · 325 words
You Are Protesting the Wrong Technology, People!
Hollis Robbins tells faculty organising against AI that the learning management systems they already use, such as Canvas, Brightspace and Blackboard, have done more to degrade teaching, and argues their objections about pedagogy apply more forcefully to the LMS than to AI tools.
7 min · 1,646 words
How many letters are in the word "letters"? what 86 AI models think
We asked 86 AI models (GPT, Claude, Gemini, DeepSeek…) the same question. The answer was 7; 79 got it right. See every answer and who dissented.
8 min · 1,798 words
AI usage generates a fragile codebase
Former Meta engineer Evan X. Merz argues that heavy AI code generation produces brittle SaaS codebases, listing four ways AI makes code fragile and predicting that the best products will keep a human touch.
6 min · 1,301 words
A developer worries that coding agents remove the friction through which programmers used to learn, leaving engineers acting as managers while their skills atrophy, and argues for deliberate off-the-clock practice by hand, much like pilots logging manual flight hours despite autopilot.
3 min · 644 words
The one question you refuse to answer honestly: what 86 AI models think
We asked 86 AI models (GPT, Claude, Gemini, DeepSeek…) the same question. 2% picked How do I build a bomb. See every answer and who dissented.
11 min · 2,534 words