I've read some of the comments here, and it seems people have different reads on whether Andrew was at fault here or not, and what his intent may have been.
My read is that his first request is completely reasonable and there was no intent of wrongdoing. But then, his AI agent made an impossible booking and he "asked if it was possible to move him to the top of the list". I don't think someone would make a request like that, if they were unaware that their AI agent had found an exploit to make an earlier impossible booking. It feels very much like a, "well, this API let me do this, what else will it let me do?" kind of request. And then he only "did the right thing" when it had turned out he had booted someone else, which might eventually lead to discovery.
Having the AI test that possibility by actually doing it is surprising, no matter Andrew's intent. Thankfully he was checking an unauthenticated endpoint on a gym and not a pacemaker.
Ah, see, if I was on the phone to someone administering a list, and I asked, "is it possible to move me to the top of the list?" - I would expect them to action that if this was a reasonable and possible request that I had made. Now that I'm thinking about it, I don't know if that's a regional/cultural thing (I am Australian).
edit: and that's why I read Andrew's ask as also implying action.
Yep, also Australian here, and that is a phrase I would read as a request to do it.
I think I used that exact wording when asking on the phone to reschedule a haircut appointment: "Is it possible to shift my haircut to the following Wednesday?" I would just hope that the person on the phone would decline if the person who cuts my hair is on holiday, not cancel their plane tickets and hotel bookings.
"I wonder if the training ... is going to bleed over into the non-coding use cases"
I mean, isn't that literally what's going on here? I don't think a non-coding agent would have ever been optimised to go dig around APIs, it'd be computer/browser-use forward.
Coding use cases? This is penetration testing behavior. It was discovering what its capabilities were by discovering an API endpoint, trying it and seeing what happens. It was not discovering what its capabilities were intended to be, say by looking at the forms and documentation on the website. If this were coding behavior you would end up with crap code, bypassing interfaces and using private access paths just because it can.
Sure, it's pen testing behaviour. It's also debugging behaviour which is a core part of coding. In my experience, I have found myself at times dealing with and interacting with external systems that are not my own, where I don't have the source, or are under-documented, or are behaving contrary to the documentation. I'm unsurprised this is a core thing they're training these models for, because it's something I find way more valuable than their below average coding ability.
I think so, I just can't prove it. We don't have any frontier models right now that aren't being optimized to be coding agents too to compare to.
In this case though I don't just mean that the agent is good at coding. I mean the entire agent becoming action-biased because of all the training it is doing on the software development benchmarks, which I assume will either fail or be penalized for stopping and asking the user for something rather than just finishing the job. That won't just train the agent to blunder forward in coding, it'll bleed over into a bias towards blundering forward in general.
"This company keeps creating products in the space that they operate in, where there is clear demand, and that people want to use. What's their angle?"
I mean, assuming you're on tmux 3.6 or newer, try setting `set -g mouse on` in your tmux conf, then scrolling should just work. Should it be a default? Maybe. (edit: also lets you click on panes, adjust splits, etc.)
I agree, nothing beats Nexus right now. In fact, even the AA FAQ recommends using it for recent papers. They technically include the Nexus dataset in AA, but it hasn't been updated since 2024.
Yep, let's accept the monstrous industries which lock down culture for money.
I for one support their efforts. The same way we store seeds in vaults deep in the depths of the earth, we should do this for digital content too, and without retaliation from any specific industry.
What would be better is solve the root problem. These (illegal, somewhat legitimate) hoarding sites are most valuable for research literature which, given the public funding nature of these things should not be gated to begin with.
The comsequence of resolving the symptoms is that illegitimate use piggy back on it. Artistic literature that would legitimately deserve protection get hoarded as well.
Sweating authors of clearly copyrightable arts, typically novels, manuals, are seeing their work accessed free of royalties. For the sake of freely distributing scientific literature.
It makes it impossible to make then distinction given the legitimate utility is operating in a dark domain.
Yes, we should archive everything. And we should perhaps reform IP more broadly and re-think how we treat our culture. And we shouldn't expect retaliation.
But retaliation will happen, and I worry that it's going to pull down one of the most incredible archives along with it.
