eddiewithzato

damn the vibecoded ones look a lot better

I can hope this will lead to little to no memory issues in using bun as a web server

That’s why I like the term “mind virus” for AI. Humans always go for shortest path

The models today will happily slop over a single 1k loc react index component on a brand new project.

They really are bad for creating a healthy codebase

Like a certain Adam Back

it’s simply that Back has nothing to gain to claim to be Satoshi. It would make bitcoin a lot more volatile. He even said just now

> I also don't know who satoshi is, and i think it is good for bitcoin that this is the case, as it helps bitcoin be viewed a new asset class, the mathematically scarce digital commodity.

That’s as close to admitting it as you can get

Look at his comment here, then you will realize why it’s definitely Back. And why he certainly would never claim to be satoshi

https://xcancel.com/adam3us/status/2041816020776611935?s=46

Also Back’s response on X was very telling

I understood minimal actions intuitively, it just made sense? The stamina meter was shrinking with each step, so I recognized it was something to look out for.

It needs to pass the most basic concept of learning, which it can’t currently do. Probably wont ever do after listening to dario on his latest podcast run.

Where we are at today is ASI (artificial semi-intelligence). Maybe in 20 years artificial super intelligence can be achieved, but certainly not AGI.

yes it’s fraud since you get money per listens

Truthfully I find sonnet-4.5 better at Rust code than Codex (medium/high). Haven't tried anything else (like react/typescript) since I only use AI for issues/problems I don't understand.

Because that's the definition that is leading to all these investments, the promise that very soon they will reach it. If Altman said plainly that LLMs will never reach that stage, there would be a lot less investment into the industry.

vibe engineering, spending so much time writing documentation for the prompt, just to spend more time debugging the slop you get.

Ill stick with writing code and just use AI for snippets/stackoverflow replacement.

Yea so basically just a better search engine. Not what the VC were promised though

Old web died when youtube reviews became more profitable.

It’s sad because I miss printed content like tech magazines.

So can many other dangerous and environmentally exploited industries. Doesn’t mean it’s a good thing

Nah it’s also environmentally, mining is bad. And if there is an alternative with no mining, people will opt for that.

Insurance company argument is always nonsense. They cannot use DNA. And if they could use DNA in the future, you would need to get your DNA sampled for the policy.

For the same reason you bought the service in the first place? They update their service with new features and update their estimates

Insurance companies cannot use it. And if insurance companies in the future would be allowed to use it, they would require you to get DNA samples for your policy.

So it’s pointless in the end

And that’s where I end up wasting more time investigating and fixing issues, rather than creating a solution ;)

I only use AI for small problems rather than let it orchestrate entire files.

You also don’t get the satisfaction of watching something grow. Teaching and being a mentor is entirely separate to massaging a prompt

LLMs will not be doing that. I wish they could, but they just spit out whatever without verifying anything. Even in Cursor which has the agent tell you to run the test script they generated to verify the output, it just says “yep seems fine to me!”.

AI at the current state in my workflow is a decent search engine and stackoverflow. But it has far greater pitfalls as OP pointed out (it just assumes the code is always 100% accurate and will “fake” API).

It’s also the worst era to play natively. Bad textures and horrible frames per second.

AI crash is gonna lead to decade long winter

Pretty sure it’s mostly for tax dodging than money laundering. Less risky too

Well you can and in fact they have narrowed down the language to a haplogroup even. R1b in the case of greek for example

Chatgpt and deepseek should be able to give you a sufficient answer ;)

Target audience is very niche