cedws

16 ms·

331.
by cedws 7mo ago
Yes, apparently OpenAI is super paranoid about 5.3 Codex being abused so if you trip their safety mechanism you have to verify your ID.
332.
by cedws 8mo ago
Claude Code’s sandboxing is a complete joke. There should be no ‘off switch.’ Sandboxing should not be opt in. It should not have full read access over the file system by default. I really want more security people to get involved in the LL
333.
by cedws 8mo ago
I let Claude trade stocks for a week and it made money! Backtesting? What’s that?
334.
by cedws 8mo ago
Could be worse. OpenAI is asking for ID verification to use Codex 5.3, through Persona, which was just exposed as doing extremely dodgy surveillance stuff.
335.
by cedws 8mo ago
Generating as much political distraction as possible so people don’t talk about MP corruption, Michelle Mone, youth unemployment, or immigration. Anything the state finds problematic for us to talk about.
336.
by cedws 8mo ago
I’m surprised that Claude assisted with this reverse engineering work. I used Codex recently for a similar purpose and got an account warning. Initially refused to do it, and then I was able to trick it. Seems I might have to make the jump
337.
by cedws 8mo ago
Macroslop
338.
by cedws 8mo ago
Somewhere along the way we forgot that technology was supposed to serve us, not the shareholders. I’m still not convinced even computing was a particularly good idea. It has enabled prosperity in some areas, like aiding drug research. It ha
339.
by cedws 8mo ago
Yeah for me it only ever polluted the context. Irrelevant information tends to oversteer the LLM and produce worse output.
340.
by cedws 8mo ago
You missed this: >which lives in the same password manager I'm not talking about Apple passkeys here, which are NOT stored in the Secure Element to my knowledge anyway. I don't see passkeys as a 2FA replacement. If they're
341.
by cedws 8mo ago
It may very well have stunted my learning. What’s the point of absorbing information when you have a consortium of experts available 24/7? Saying what you said about it being down to being how you use LLM comes from a privileged positi
342.
by cedws 8mo ago
We should be very concerned for the next generation. When you have the constant temptation of digging yourself out of a problem just by asking an LLM how will you ever learn anything? My biggest lessons were from hours of pain and toil, sco
343.
by cedws 8mo ago
We have Copilot and Claude Code. The rationale of having both is that Copilot is cheap anyway, grants access to a range of models, and some developers still haven’t moved over to CC. There’s also the smart auto completion which is actually
344.
by cedws 8mo ago
Next week Anthropic will do something evil and everyone will be moving back to OpenAI. Crazy thought but maybe we should regulate AI instead of relying on the hegemony of three companies to police themselves.
345.
by cedws 8mo ago
It doesn’t have to be genuine concern, it may just be his internal narrative. He allegedly raped his own sister. No charges have been brought against him.
346.
by cedws 8mo ago
There’s another foot gun I wrote about recently: https://cedwards.xyz/passkeys-are-not-2fa/
347.
by cedws 8mo ago
Excellent post.
348.
by cedws 8mo ago
I think Altman probably rationalised it to himself by thinking that if he doesn’t do it, Musk/xAI will, and they give zero fucks about safety. So maybe he told himself that it’s better if OpenAI does it.
349.
by cedws 8mo ago
The billion engineers building sandbox tools at the moment are missing the point. Sandboxing doesn't matter when the LLM is vulnerable to prompt injection. Every MCP server you install, every webpage it fetches, every file it reads is
350.
by cedws 8mo ago
In recent years I’ve noticed a massive rise in what could be called ‘financial degeneracy.’ Shitcoins, NFTs, sports betting, even stuff like Pokemon cards just seem to be a vehicle for trying to make a quick buck and find a bigger idiot. Pr
351.
by cedws 8mo ago
People say you don’t need Redis, you can just use Postgres LISTEN/NOTIFY. I read a blog post about it locking the entire database and causing problems[0]. Apparently it’s fixed now but this only just being discovered in 2025 doesn’t gi
352.
by cedws 8mo ago
Yes, I’m more worried about the false confidence such technology could create. Implement an authenticity mechanism and it will be treated as truth. Powerful people will have the means to spoof photographic evidence.
353.
by cedws 8mo ago
How much are these price hikes affecting the ability for people in poor countries to access compute?
354.
by cedws 8mo ago
I really hope so. I moved to Codex, only to get my account flagged and my requests downgraded to 5.2 because of some "safety" thing. Now OpenAI demands I hand my ID over to Persona, the incredibly dodgy US surveillance company Dis
355.
by cedws 8mo ago
Like an abusive spouse. 'No, you can't leave me, you need me.' Actually, we don't. We used to have a good relationship and you lit it on fire. Bye, US.
356.
by cedws 8mo ago
We’re off to a great start then with Anthropic banning users who use alternative clients with their Claude subscription.
357.
by cedws 8mo ago
I’ve always wondered what made him turn around. Threatened his family? Was Prigozhin a family man?
358.
by cedws 8mo ago
Governments in Europe should be seriously scrutinising this with the background conversation of departing American tech going on. Discord users globally were being coerced into handing over their ID to this American surveillance tech. Are w
359.
by cedws 8mo ago
There’s still sane people out there, I’m one of them, watching this gigantic trash heap ready to go up in flames. It’s not just OpenClaw either, it’s everything. Nobody is paying any attention and when it goes wrong it’s going to be an abso
360.
by cedws 8mo ago
I've also had issues topping up my (virtual) Suica card late at night before.

More ›