AI Enthusiast Log

Sign in Subscribe

Topic

llm

A collection of 34 issues

AI agents’ token usage surge makes seat‑based SaaS pricing untenable

DeepSeek's 75% price cut on its V4-Pro model makes inference cheaper at the per-token level, but the source frames this as insufficient against a structural problem: agent workflows amplify token usage from roughly 1:5 in a single-turn chatbot to 1:700 or higher in

Slopsquatting frames LLM hallucinated packages a supply‑chain attack vector

Slopsquatting reframes LLM hallucinations as a supply chain attack surface rather than a quality problem, according to a recent ReHack analysis. The technique works because hallucinated package names are statistically predictable outputs, not random errors. That predictability is the property attackers exploit, because any name a model generates at scale

Credit card payment of $67.53 for a self-pickup order with a $66.85 discount, completed on 2023-12-26.

A new study evaluating 67 frontier models from 21 providers has measured something the multi-model orchestration approach has long assumed away: a shared failure mode that the standard diversity metric cannot detect. The paper's authors call it the co-failure ceiling, and on the MATH-500 benchmark

Morgan Stanley's FIXR cuts P&L time by making agents less autonomous

Morgan Stanley's FIXR system cut P&L reconciliation work roughly in half, saving about 1,500 hours a week across roughly 100 controllers. The result came from an architecture that runs against the dominant agentic-AI narrative: the team designed the system to be less autonomous over

Anthropic prices Claude Sonnet 5 for IPO-era adoption, not capability gaps

Anthropic released Claude Sonnet 5 with introductory API pricing of $2 per million input tokens and $10 per million output tokens through August 31, rising to $3 and $15 after that. Standard Opus 4.8 pricing is $5 and $25. The capability story the company wants to tell is that

Google's NB2 Lite undercuts NB1 on price, restricted to 1k resolution

Google is shipping Nano Banana 2 Lite, technically Gemini 3.1 Flash-Lite Image, at $0.034 per 1,000 images with a 4-second generation target, framing it as a low-cost, high-throughput utility for enterprise workflows. According to internal documentation cited in the source, the price undercuts

Agentic AI deployment is outpacing the data needed to train it well

MIT Sloan and BCG reported in November 2025 that 35 percent of surveyed businesses had already deployed AI agents, with another 44 percent planning to do so. That 79 percent share of enterprises moving toward agentic AI is the headline. The less visible story, surfaced in an MIT News Q&

LongCat-2.0 reveal: a narrow coding win backed by a caching play

Meituan confirmed that the OpenRouter-leading stealth model "Owl Alpha" is its newly open-sourced LongCat-2.0, a 1.6-trillion-parameter Mixture-of-Experts system that posts a narrow 59.5 on SWE-bench Pro against GPT-5.5's 58.6. The release lands

OpenAI updates GPT-5.5 Instant, but chat-latest is not the API model

OpenAI updated GPT-5.5 Instant on June 24, 2026, less than two months after the model's spring debut, with the company claiming better intent recognition, more reliable handling of complex multi-part instructions, and stronger shopping and local recommendations. The company has not published benchmarks for the

Self-Harness automates agent harness tuning, but evaluation gates the gains

Shanghai Artificial Intelligence Laboratory researchers introduced Self-Harness, a framework that lets LLM-based agents rewrite their own operating rules by mining execution traces for failure patterns and proposing targeted harness edits. On Terminal-Bench-2.0, three model configurations improved 33 to 60 percent relative to baseline, with the