People complain about AI slop – but it is really good at generating unsettling concept pieces such as this one. There’s even courses on how to make these kinds of shorts.
“You were always a rounding error in the equation of eternity”
In just the last 6 months, LLM’s are now find bugs and security issues in code that nobody has found in literally decades of investigations.
More terrifying? This isn’t your game code – it’s code that runs everything. Bugs in the Linux kernel that have been there since 2003 but were found by an LLM in minutes. Bugs in packages that have been around for 25 years. Bugs so bad that they give you access to almost anything. The rate of bug discovery by LLM’s is now on an exponential curve that is unlikely to stop anytime soon. In fact, he knows of hundreds of kernel bugs he cannot file because he simply can’t keep up with verifying them at the rate the LLM’s are finding them.
Based on current AI progression, he believes LLM’s will be better at finding bugs than the top 1-2% of vulnerability researchers in less than a year – and that power will likely be available not with GPU’s in the cloud, but on LLM’s you can run on your laptop. With this facing the world of computing, it’s highly likely bad actors could nearly completely destroy and hijack banking systems, just about any online service, and critical infrastructure. Almost everything could be open to hackers in just months.
We’re facing a real Y2K situation. It’s happened so fast that almost everyone is in denial or simply doesn’t know it’s happening.
kaiyimi has one of the better videos on getting an SGI IRIX system. Looks like you have to use specific versions of MAME64 (223) and jump through lots of hoops – but it seems like a workable solution.
I would love to know how he set up the original hard drive image.
If you’re looking for me, I’m with all the grown men in the corn.
I didn’t even realize the old Bop-It toy had an ending. The original edition ended after a maximum score of 100. The second edition went up to 200 points.
This song is amazing.
The internet is just now becoming aware that many new cars have cameras that watch you. Everyone is ascribing this to insurance companies, government control, and so forth.
The reality? This was mandated under the Biden administration and the EU for all new cars to have drowsiness detection systems in all cars. One of the solutions car manufacturers have found is to point a bunch of cameras at you and do gaze detection.
If people are freaking out about this, wait until they hear that all cars and trucks starting in 2026 must have passive drunk driving testing systems before you can start the car – a law which is ironically being defended by alcohol companies. Can you guess why?
Yosh has been covering the increasing use of AI to test for better and faster racing times for track A01. The conclusion: AI was able to bet the human world record, but is still lagging the hand-tuned TAS (tool assisted speedrun). They’re not done trying yet – we’ll have to watch and see how many more records fall to AI and which do not.
Just a few months ago in 2025, the curl team was getting inundated with an explosion of AI slop bug reports. They went from a real issue rate of about 15% to less than 5% of reports being real issues. Many were almost nonsensical submissions by AI – in attempts to gather bug bounties.
That has all changed.
Daniel Stenberg is now reporting that they’re getting really good security reports – almost all done with the help of AI. It’s gotten so many, they’re having trouble keeping up. Combine that with what is coming with systems like Claude Mythos that are finding thousands of zero day bugs in every OS and browser, and the software world is likely about to see a massive software security upheaval that will leave the development world dazed.
The revolution that is coming at all levels of software is clearly starting to rival what happened in the dot-com era.
I also heard this from another person that worked directly with Steve Jobs. Fascinating comparison with Elon Musk as well.