By the written contract, they cannot use the data for traning our company gives them. (according to our lawpersonel, I did not read the contract)

Also developers had to sign additional NDA to be able to read the specs. And it is strictly forbidden to give AI access to restricted documents. Yet it learned somehow. That is stealing.

I have to agree, the recent jump from shitty as best to “oh fuck i will be replaced” was due every fking company jumped on the hype train and shared everything, and they even pay for it.

I work as an embedded developer as well. I would have agreed with you some months ago. Now it can recognise a full proprietary compiler by one of its flag, which is used maybe a handful of companies. Scarry. I am 99% sure it “learned” by previous attempts by developers. Aka the AI company stole the conversations and trained the new model on them. The uC we are using is non public, only we have it, yet somehow the AI knows about it. The only thing it cannot do is grabbing an oscilloscope to debug the HW…

This argument is flawed, as I just took a break from reading disassembly to scroll a bit. Assembly is the human readable representation of machine code. It is relevant today (eg embedded development).

Also each move to higher level comes with its own set of issues and trade offs. I think you somewhat correct to correlate with the past advancements, but LLMs fundamentally different: It is not a new higher language, its normal language. Also it cannot grab an oscilloscope and debug HW issues which are wrongly documented by the supplier.

Uff, thats really tough. Hang in there! I was fortunate enough to land a job, mostly because ppl I knew from uni recommended me.

Jeah, I am pretty disapointed that AI learned to code on my projects. Sometimes I wonder if it poisoned the model a bit at least… I had everything published I ever written, should have kept them on the floppy.

Bold you to assume there are sufficient client specs. Maybe the monetization path (yes to all manipulative way to squeeze money) ready.

I enjoy Kurzgesagt videos. They are a Science communication platfrom. And they simplify things greatly. So with literally every other Science communication, the goal is to show a topic and the audience should research it further. Anyone who thinks they provide whole objective truth is a fool. This does not make them evil, nor not evil.

I tried to watch the rumblings of the provided video, but I was unable. However I read the metr report

I highly recommend. Even for the non-tech savy. From page 5 to 20 it is a greate sci-fi story. Keep in mind tho: OpenAI provided the data, they used OpenAI’s AI to structure/mine the data, and OpenAI had 100% access to redicate anyting.

My conclusion: Fuck AI companies. Especially OpenAI where incompetence is so high I might think it was deliberate.

I cannot fanthom they have 0 observability on a fking shared resource (artifactory). Why it was shared. Why agents had write permission in the 1st place (at least to create directories)

Why there was 0 monitoring on agents. I can’t beleive they let programs run for days and hav 0 idea whats happening.

Why they were not separated from the internet, they had offline task.

Why impossible tasks was given to agents. Why the flags were generateble (default seed) Why agents knew they are tested? Etc.

This reeks of Incompetence or deliberate Malice. Malice not against running program code, against humanity, by creating an environment where the only possible solution generates a huge uproar and makes more money.

Also if a actual human would have done this, they would be sitting in a nice chilly jail cell. Why there is no responsibily? I really don’t like to be a precedent of “woops my AI did it, it was a honest mistake I promise”