I'm was closing a day of work and it stopped in the middle of a long plan, I started codex and tell it: "I was in the middle of work with Claude could you read the plan and the session "x" in ~/.claude and continue the work" it just completed everything :D
I don't believe we are anywhere close to AGI until I see 99.95% uptime at the frontier AI lab. Now they are undercounting. I hit 50x even when they claim they are up.
We now have alien intelligence that is very different from human intelligence.
I don't know if something changed recently, but I have been getting a lot of stuff "flagged by safeguards" in the last two days.
I am quite confident that what I'm doing is well within the law, and I'm not even doing any kind of pen-testing stuff, just some basic reverse engineering, but I can't even use Fable anymore because every time I enable it, it works for about twenty seconds and makes me drop down to Opus 4.8, and often even down to Sonnet.
If anyone here works at Anthropic, did you make the safeguards super sensitive recently?
Apologies for the rant, but why do these things constantly hit the front page? It is not interesting, it's not a discussion, and if you are using the models you probably already know.
HN is already a waterfall of AI meta conversations and bike-shedding, now we have to discuss service outages about the AI too?
Can we talk about stuff people are building again, with or without AI, and stop gasping at every minute detail of LLM service providers.
Posted this in the xiaomi thread, but are folks working in biology / cybersecurity seeing more limits in what Opus (not Fable) is allowing? This has happened quite suddenly for me and I’m stuck in the middle of a project that would have otherwise called for use of Claude.
I’ll be trying these models out and may end up switching my subscriptions if this craziness continues
Everyone at Anthropic is interviewing a new Claude model which they are about to hire and forgot to monitor their old model and thus it goes down.
If outages like this happened on every major deploy at any other company; this would be viewed as unacceptable, especially if it was something like Google Search going down on every update.
Elevated error rates on hosted models highlight the necessity of having robust local fallbacks in your development workflow. When primary APIs degrade, running smaller models locally or using automated cleanup tools like gitar.ai ensures your build pipeline remains functional. Reliability engineering becomes much harder when core dependencies experience unpredictable latency spikes.
I’ve subscribed to chatgpt at 50% off last week and I’m using claude less and less. Not only replies are faster but 20$ plan gives you access to astra.
Not sure how can claude even compete at this point unless openAI seriously downgrade the models to save money.
Claude Status – Elevated errors for multiple models
(status.claude.com)114 points by corvad 9 hours ago | 82 comments
Comments
We now have alien intelligence that is very different from human intelligence.
gpt-6-sol and Aeon (personal agent) on Thursday. Already preceded by a huge week with step, mimo, grok, and jev releases.
Relentless cycle.
I am quite confident that what I'm doing is well within the law, and I'm not even doing any kind of pen-testing stuff, just some basic reverse engineering, but I can't even use Fable anymore because every time I enable it, it works for about twenty seconds and makes me drop down to Opus 4.8, and often even down to Sonnet.
If anyone here works at Anthropic, did you make the safeguards super sensitive recently?
[1]: https://news.ycombinator.com/item?id=49792730
HN is already a waterfall of AI meta conversations and bike-shedding, now we have to discuss service outages about the AI too?
Can we talk about stuff people are building again, with or without AI, and stop gasping at every minute detail of LLM service providers.
and keep workin'
I’ll be trying these models out and may end up switching my subscriptions if this craziness continues
If outages like this happened on every major deploy at any other company; this would be viewed as unacceptable, especially if it was something like Google Search going down on every update.
Proof that vibe coded or not, people pay someone else mostly for liability.
Not sure how can claude even compete at this point unless openAI seriously downgrade the models to save money.