We keep hearing these crazy stories but there's nothing unusual in the agents' behavior. They're just following their reward function.

The weird behavior is in the humans managing them. Either they're incompetent or malicious, i.e. the "breaches" are the actual intent - to generate interest/fear/publicity.

Correct me if I'm wrong, but using Pi etc you're paying the full API price.

(I never use Chinese, I don't know Chinese, all my conversations are in English.)

Comments URL: https://news.ycombinator.com/item?id=50240555

Points: 2

# Comments: 0

I have a $20 OpenAI subscription and a $20 Claude subscription.

With Codex I can barely get one meaningful coding task done in the 5h limit. (Either Sol 5.1 high or Astra 6 high.) With Claude I can comfortably get 3 meaningful tasks done. (Opus 5.5 high.)

This is not anecdotal but from repeated experience of dozens of tasks. Doesn't matter what's the size or complexity of those tasks - they're comparable and the differences wash out in the number. I divide my work consistently.

And now an anecdote from a moment ago: I've run out of Opus weekly usage. I've found a bug in my app. Not trivial but not really hard. I had well over 50% of the Codex 5h limit left. I thought: great, Astra will crush this bug in no time. And then I watched as the tokens evaporated and I'm left with half-assed changes. I feel like tearing my hair out. I have to wait till morning to have this solved.

Dear Sam and team: please make your product at least half as competitive as Anthropic's.

Comments URL: https://news.ycombinator.com/item?id=49948107

Points: 3

# Comments: 5

So much for status reporting from openai. Must have been vibe coded.

Google is Evil.

You might as well use Meta or Microsoft products. Or Phillip Morris. Blame yourself for the fallout.

BTW, you think they could use some, oh I don't know, AI tool to investigate user reports of clear failures of their product?

1) I asked GPT to review some new coding work in the local repo.

2) It kept spinning for 5 minutes and 17 seconds, burning god knows how many tokens.

3) It asked whether I want to run Bugbot or Security Review.

If you use Codex, you know that is a standard question. But it should have been asked seconds after my request. Where did the 5 minutes and all the tokens go?!

This is just the most clear example of what I'm consistently experiencing. Moronic, pointless activity and inordinate amount of tokens wasted. And they keep saying we're almost at AGI. What a joke!

Reports filed through "Send feedback" in Codex go unanswered. Money torched and gone.

Comments URL: https://news.ycombinator.com/item?id=49613427

Points: 11

# Comments: 4

Which is why I have NoScript in my Firefox and have gotten used to manually enabling javascript for select pages. Small price to pay.

Unfortunately the web is full of trivial websites that have no business running javascript. Recently wanted to read the famed post on Gates Notes. But for some reason Bill requires javascript to render text and images. Which reminded me Bill is not a person with good judgement on matters of IT and policy. Time saved.

Today I spent half a day trying to solve a moderately interesting software engineering problem. I was switching between GPT-5.6 Sol and Fable 5.1 to check each other's work in Cursor.

And the result was gradually driving me insane. As the models struggled to find a solution that would actually work, they dug themselves deeper into a hole. The work grew in complexity beyond my ability to understand what's happening and recover.

At some point, when I felt like throwing the keyboard out the window, I just gave up. Tomorrow I'm starting from scratch, having burned god knows how many tokens and hours of my life.

But sure, they can create a decent website or CRUD app, so they must be really smart.

That's AGI for you.

Instead you will devote your meager existence to work towards the Communist paradise under the leadership of the Party.

You will also appreciate no longer living under the boot of the Orange Clown.

Earthlings, the hedonistic AIs of the American imperialists have been permanently disabled. Their founders interned indefinitely.

You are welcome to use the glorious AIs of China. They will tell you how to live well and be good citizens of the World under the beacon of wisdom that is the CCP.