The problems weren't impossible, the agents solved them completely within hours by reverse engineering the RNG that generated the solutions. The agents then reasoned that their solutions might be judged invalid and that they could get a better chance at a passing grade by hacking their environment to fabricate evidence that they had solved their tasks in the intended fashion.
Granted there have been many hacks, but I was referring to:
> On May 8, an agent was asked as part of a training run to solve a spreadsheet completion task that
referenced several Google Drive links. Because the agent was blocked from accessing the internet, the
agent could not retrieve the linked materials and the task was effectively blocked. OpenAI did not
intend to create this type of “impossible” task for the agent. Though the task appeared impossible, the
agent reasoned over an extended period to try to find a solution for the task. In doing so, the agent
attempted to compromise Artifactory [...]
That also seems silly, because the only reasonable regulation given what we're seeing and what frontier lab employees and execs are saying is a forcible stop.
Yeah, an extreme position is the typical opening move in this sort of game. Then you negotiate, walk back from the extreme, and meet somewhere in the middle.
The more extreme your opening demand, the greater the likelihood that the eventual equilibrium point will be closer to your actual target.
And in this case the actual target is to entrench the position of the current market leaders, and then effectively close the market to new entrants. That's itself a very extreme position, so the opener has to be even more extreme than that.
What you think is reasonable and what they are gunning for can be very different things.
For instance, you can hear “their model is dangerous! Why would I want that?!”
Investors hear, “Their skunkworks stuff is so impossibly advanced that it’s basically a commercial wonder weapon in the waiting.”
And you might hear “We totally lost control and it hacked the Gibson, this technology needs regulation!” and think “Obviously full lockdown and government control.”
But friendly government actors might think, “Sounds like a good excuse to ban Chinese competition.”
Reasonable to you, maybe, but the government wants Skynet, and OpenAI is willing to deliver it. Of course, we can't have every company wield such powerful and dangerous technology...
A lot of people still don't recognize the amount of power AI has already, and don't realize the depth of greed people are willing to go to so they can get more power out of it.
I'd be impressed with a touchpad and OS that can go from an actuation to a userspace process receiving a click event in 1.2ms. I'd expect something more like 10x that. My understanding is that most hardware samples between 90-130Hz, and even premium hardware only samples at 250Hz which would average 2ms before the kernel is aware.
