hoppp

I don't know anyone running Paint servers and I'm pretty sure phishing could be possible.

It's probably a way to get phished, if you click on a malicious bmp file.

End year sales are common. Black friday, Christmas whatever. People spend and taking profit feels good.

Was this exploited in the wild?

On android they fall down along the side all the way to the bottom

There are always more left.

Well at least they are doing public service while hacking the planet.

Go is lovely. Best language to generate backend services with.

I like it. I never really liked marshalling anyways, It was weird.

I'm happy it's getting fixed!

"Disaster has been averted"

Yeah, hmm a lot of people don't actually auto update their PHP so.. it's not like the vulnerability disappeared from the wild.

They can tokenize AI tokens to speculate on futures or offer options

My approach has been that I architect a system myself that is modular and the agent tackles one module at a time. If the module does not work it can be ripped out and replaced without effecting the rest of the working modules due to strong decoupling. This way I may ship code I don't audit but when I have to the issues are already easy to find. And of course document and unit test everything.

So I don't cognitively surrender the system design, only the inner implementation of the modules.

Blockchain dapps were getting hacked long before AI. Not a signal

I think we did good inventing it and it's inevitable. Lets use it well.

Yes, in everything.

I stopped using my laptop for dev and switched to talking to claude code on my phone and as you can imagine I am faster at producing code.

100% cloud with a single entry point device to manage dedicated servers + agents

If we can plug in a boolean circuit, I wonder what else we can do. What about plugging in something that can serve as dynamic long term memory?

Do not use frontier labs products if you want to avoid leaking IP

host your own models. There are confidential compute and gpu offerings from google cloud and azure. Pay as you go, rent your own GPU and host your own models there.

Maybe thats gonna be the way to make LLMs self improve better?

Similar to implicit learning in humans when the subconscious creates new heuristics based on observations, one LLM could adjust the weights of another to tune it to the environment

If the entire project is developed with AI it can add features effectively but for large old codebases it becomes an issue

I am not very against data centers, Im just trying to understand why people talk about water consumption more.

Water is something to protest against now because it's very visible. There is a daily water consumption.

The climate impact, like with cars or the oil industry is a boring future issue.

People have always been bad at critical thinking, it's a high level cognitive skill that requires practice.

Now that it can be outsourced,we will slowly lose this ability that we barely had.

Generally we don't hear much about attacks in Java like that. But they probably exist but python and javascript are targeted mainly because everyone uses them. They cover a lot of ground.

I am now developing with claude code on mobile and sandboxes and a few remote VPS I rent.

No more laptops, just a dedicated android device.

I developed a devops/admin tool and then its claude code + github

They use open router.

I've seen attackers install claude code on compromised servers

Yup, it's quite fascinating to watch.

Yeah, why is there $6bn in hot wallets in the first place ..

Makes me think traditional banks have solved this from the start. You can't steal the backing of an unbacked currency.

Yeah, 9 shares and then 10k people click

I already sucked at incident handling before, it was more of a panic attack mixed with the fear of getting fires. So Agents have a good place there.

It's hard coded probably.