Claude’s new model is more ‘honest’ when it messes up
… Higher-effort responses will use more tokens, giving users the option of lower-effort responses if they don’t want to burn through their rate limits as quickly. …
Search
All sources
xda-developers.com
25
anthropic.com
7
androidauthority.com
4
tomsguide.com
2
theverge.com
1
macrumors.com
1
androidpolice.com
1
9to5mac.com
1
wired.com
1
theregister.com
1
bleepingcomputer.com
1
Anthropic Launches Claude Opus 4.8 With Gains in Coding and Honesty
… Testers have found Opus 4.8 to be "more reliable and sharper in its judgement" when doing agentic tasks, and the model also made gains in honesty. Early testers report that Opus 4.8 is more likely to flag uncertainties about its work and less likely to make unsupported claims. …
Claude Opus 4.8 is more honest, less deceptive, and considerably cheaper
… Learn more. General technology Claude Opus 4.8 is more honest, less deceptive, and considerably cheaper The company is also adding a couple of new features to Claude. …
Claude Opus 4.8 just proved AI is finally growing a backbone — and it crushed ChatGPT in 7 brutal tests
… The model completely refused to flatter and exposed the self-delusion with intellectual honesty. …
Claude Code with a local LLM running offline is the hybrid setup I didn't know I needed
… Anthropic developed Constitutional AI CAI , a technique that trains Claude using a set of principles — a 'constitution' — to guide its responses toward being helpful, harmless, and honest. …
I replaced the expensive Claude Pro subscription with these local models, and my productivity didn’t drop a bit
… Anthropic developed Constitutional AI CAI , a technique that trains Claude using a set of principles — a 'constitution' — to guide its responses toward being helpful, harmless, and honest. …
I made Claude slower and it completely changed how I use it
… I started questioning the process more. In one test, the response was around 300 words, and it took more than 2 minutes. Normally, I’d read it in 80–90 seconds, so this was noticeably slower than my natural reading pace. Yet I engaged more. I wasn’t skimming anymore; I was following. …
I ditched Claude Pro for free tools for a week — and one of them had no right being this good
… Anthropic developed Constitutional AI CAI , a technique that trains Claude using a set of principles — a 'constitution' — to guide its responses toward being helpful, harmless, and honest. …
Claude Code finally remembers why I made those choices, and my workflow is faster because of it
… A tool that remembers more also creates more responsibility. The privacy concern isn’t theoretical. Development sessions can include tokens, local paths, API responses, private project names, and other details that were never meant to become durable notes. …
I used Claude Code, Google Antigravity, and Codex for a month and I have a clear winner for you
… Anthropic developed Constitutional AI CAI , a technique that trains Claude using a set of principles — a 'constitution' — to guide its responses toward being helpful, harmless, and honest. …