Is there any way to have frontier models solve captchas? Im so tired of those just breaking my workflow
I have external tools that they could use to solve it, but they just dont want to. Like, webpages could just add captchas without the annoying challenge for humans and agents would stop anyways
Codex error event: Codex is at capacity, please try again later (code=turn_rate_limit_exceeded)Codex error event: Codex is at capacity, please try again later (code=turn_rate_limit_exceeded)Codex error event: Codex is at capacity, please try again later (code=turn_rate_limit_exceeded)Codex error event: Codex is at capacity, please try again later (code=turn_rate_limit_exceeded)Codex error event: Codex is at capacity, please try again later (code=turn_rate_limit_exceeded)Codex error event: Codex is at capacity, please try again later (code=turn_rate_limit_exceeded)
μck ٤: retweeted
bunnies i guess has been placed at #1545, above mischievousmechanism and below Crimson Citadel.
This pushes Precipitance out of the top 1600.
I just tried taste-skill since I needed to make a shitty-but-not-so-shitty website and im heartbroken by how solid it is
something something first 90% is easy last 10% is hard
been building an iOS app for the last 4 days and it's pretty much done already. I just have to wait for some bureau stuff and for the App Store review
though that was the easy part, now it's time for marketing (god kill me please)
if this makes me any amount of money/profit I have definitely been putting my focus on the wrong stuff lol
if OpenAI next release isnt more usage-efficient and better than Fable 5.5/Opus 5.5, theyre out of the race till they somehow figure out (and fix) what has been going on lately
they patched the method / error lol
if you were a verified Student, you could upgrate to Pro (from Plus, the one youre assigned for free) for 5$, and it wouldnt charge you again until next year
I had my upgrade removed, and it seems like it doesnt work anymore T_T
first image now vs second image 2 days ago
Deepseek the Cybersec Demon.👀
We found DeepSeek V4.1 Flash 1017 beats GPT 6 Sol on our multi-step offensive cybersecurity task. Sol was tested via the Daybreak trusted access program.
We have been building a cybersecurity benchmark based on simulated corporate networks. This is a valuable indicator for the cyber-risk a model poses if in the hands of a bad actor. We plan to make the full evaluation public in the coming weeks but wanted to share our first immediate finding here.
been saying this
Tim Sweeney is criticizing Valve again over Steam’s requirement for developers to disclose their use of generative AI.
The Epic Games CEO called the policy “abhorrent,” arguing that developers are being forced to make a “political statement” that could be used against them by a “hate mob.”
Sweeney has been criticizing the policy since 2025, previously comparing the disclosure to a “Scarlet Letter” and arguing that AI will eventually be part of almost every game production process.
scarlet letter is a neat name
Replying to @JustMicrock
modrinth is a good platform (much better than curseforge has ever been) but its sad how they are trying to appeal to the anti-AI crowd by adding a (useless) feature that will, undoubtedly, only cause harassment for some mod makers
because thats the only purpose of this mark-as-AI label that platforms are introducing :P
people who dont care about AI will simply ignore it, and people who do will hate on it for no reason whatsoever other than it being marked as such, as has happened many times before and will happen many times in the future


























