pixel_popping

That's how I feel about it, productivity gain of 10-100x is definitely real, but switching from programming to being a callcenter agent is hard on the mind, I know personally many people depressed because of it, the loss of ability to write code is also quite real, it's not clear if it matters for the future, but it's sad to see decades of problem solving and software engineering turning into speaking with a machine all day long.

AI has never been as cheap as now, can you elaborate on the "bubble is bursting"?

Insecure human slop on top of that. Projects not checked in-depth by AI is recklessness at this point, prove me wrong with an example.

I've found 2 serious vulnerabilities in Librewolf (can crash computer from HTML page, from an iframe) as an example this week, as per their rule, agent isn't allowed to report it, guess we will just have to leave those critical around or wait around that a human discover them while attackers already exploit them in the wild :/

Codeberg doesn't want the world to be secure, quite shady, unless they care to explain what's the downside of discovering critical vulnerabilities with AI, especially automatically? Thinking that humans will write more secure programs than machines is just wrong and shows an outdated and dangerous mindset.

Humans+AI > Humans.

Opus 5.5 is currently cheaper than Deepseek V4.1 flash (with APIfied subs, but who cares, that's the price for individuals & startups, unsure why everybody seems obsessed about "API is the price" while it's genuinely not except for Enterprises specifically) - It's quite hard to justify switching outside of the privacy factor.

For us who spends more than $150K tokens a month with less than $5K spending in subs, it's really hard to justify the switch and it's even senseless (for now).

Circular deals have always been part of the top companies operations and is absolutely known from a financial perspective since forever, nothing new.

I believe that will continue regardless of regulations, as long as Co Y and Co X is allowed to buy from Co N, then that can't be fixed.

Great work honestly. You could genuinely expand this into a worldwide matter, there might be a momentum to grab now.

> We recommend investors avoid this IPO based on the company’s: warning, in its own prospectus, that is poses “a catastrophic or existential risk to humanity”,

Why would this matter for an investor?

Yeah, it runs Flash, RE to Rust with a WASM backend inside a v86 VM, standard in late 2026.

Remember when they put that 25GB daily limit vs MU unlimited, that was the start of its death.

It's such a trashy deceptive blanket statement "has delivered zero financial payoff to corporate America outside of tech." - literally I know at least a few companies myself that drastically benefit from it and have lowered their operational cost by a lot.

That's just a false statement.

Back to prompting now.

Why wouldn't they do that already, why would they refuse to process B$ equivalent of data? The data are in cleartext in their DB and no one can verify, I would come from the assumption that it's already the case, their current ToS are irrelevant.

Why speaking in Tokens as it's the only "AI currency" available while completely discarding that there is no problem having 50 subscriptions mixed for a startup?

Pretty fun! thank you

In FAQ:

> Only the computer answering it, and it isn't kept afterwards.

Why wouldn't it be kept afterward, there is absolutely no guarantee that the owner of that node won't just retain everything? So basically any rogue GPU node can inject tool calls in responses or just collect prompts...

Are you using some sort of remote attestation/enclave for this?

It's totally realistic, knowing I've found multiple CVE myself doing exactly this method 5000x (we do loops that way almost everyday) in very serious softwares including Bun, Node, LXD, Openresty and so-on.

Why not do 1 more prompt to make the website trustable?

It more reflect the fact that programming without deep AI verification is reckless today.

Logo is blurry fyi.

It's now rolling on the k8s clusters, thank you for the great contribution.

Very disappointing to add trackers within that page (to Cloudflare insights) which virtually allow correlation of those users and who uses Monero and that's accessible by the US Gov :/

getmonero you could do better.

It's not true, fingerprinting on Safari is also easy regardless of Incognito, VPN and whatnot.

Different computers doesn't solve a lot of fingerprinting methods as well.

Fingerprints are used for over a decade already, nothing new, just users keep closing their eyes despite knowing, even advanced ones.

Error: Verification failed. Please refresh the page and try again. (Captcha passed successfully) :/

What's the actual difference with Openrouter or Eden? They propose all those models I believe.

Altman or OpenAI? Weird inaccurate headline, I haven't heard CEOs getting personal fines?

When Apple got hundreds of fines because of illegal behavior, we didn't generally see `Tim Cook`, strange?

I hate saying this myself (I'm an immigrant from EU) but the vision that robots will be great enough to actually solve the problem in part is quite realistic imo, very realistically, in 10-20 years, humanoids will be everywhere and probably very competent, even more so than most humans (or at least "young" humans that are inexperienced in a general sense). It's sad, but it's viable.

Truthfully, it was an AI slop and I don't even know how Mozilla couldn't see that, it didn't even make sense commercially speaking knowing that close to all competitors were doing better already.

It seemed like the "plan" wasn't even really thought for the next year or two, it was just some sort of experiment which is useless commercially speaking, no one expect "experiments" from major companies like Mozilla, they expect real products that is competitive and works for the years to come.

There is no "experiment" to make anymore, era of MVP is mostly dead since years, only well-thought and durable projects will succeed, getting a few thousand signup doesn't prove anything anymore, it just proves that those customers just weren't aware of the more advanced competition thus will get swallowed anytime soon.

This page couldn’t load

A server error occurred. Reload to try again.

error "Tool not found"