It’s not “endless complaining” when the bot is functionally cheating. Yes if you let a computer use special mechanisms it will operate faster than a person who is using a mouse and a keyboard. A computer computers better than a person, what a revelation.
I don’t applaud people using aim assist in CS for the same reason. We literally don’t have access to the same tools. That’s why chess has been such a great target for years. One’s ability to manipulate the pieces has no bearing.
Work in support for most of your life and you come to the conclusion that there are a whole lot of meat proxies out there. In fact everyone does the meat proxy somewhere in their life.
> I started seeing people turn off their brain as they use LLMs
I mean, this was true for computers. "Computer says no"[0] was thing long, long before LLMs. I've had well educated, seemingly very intelligent people break when they couldn't log in steadfastly ignoring the [caps lock key is on] error message. This becomes even more "fun" and "interesting" when a person is being a meat proxy for some horrifically dangerous industrial process involving high pressures, excited chemicals, and enough rotational inertia to send god flying.
China is much more cagey about the exact levels and capacity of it's version of the SPR but there's a lot of evidence they're drawing pretty heavily from it to cut their oil imports which has had a big impact on reducing the impact of the whole illegal war in Iran situation.
Prisoner’s dilemma: margins are razor thin. If your fellow farmers use better seeds then you are forced to use them as well. Otherwise you won’t be able to match their prices without selling at a loss.
That's true of any productivity enhancement. If your neighbors use more efficient tractors you're at the same disadvantage, but nobody argues that farmers should be able to steal tractors.
The parent has to do with permission, not stealing. Imagine if your neighbors could have an exclusive license to buy the new obviously superior tractor which you will never be able to obtain no matter what you do, simply because you're not them. This is legal today, the argument is that it shouldn't be because it punishes competition and lets the rights-holder wield too much power over the market.
Only if the cost of monsanto seed + roundup is less than the cost of regular seed per expected harvested bushel? Otherwise, no one would use the monsanto seed + roundup?
One, the move to digital lives made digital surveillance necessary. It wasn't enough to check attendance at the local mosque/far right rally. Then, the counter-move to encrypted communication meant law enforcement now had to rely on metadata.
Second: A wanton disregard of the treasure that is digital data. Most people just don't seem to care. LG is snooping on you? I bet they won't see even a 10% drop in sales next year. When the citizenry as consumer blasts their data everywhere, it's hard to argue as parliamentarian that those same people want it protected. Now people use Apple Watches and Amazon Alexa in their own bedrooms! People say they want their data safe guarded, but their actions suggest otherwise.
> A wanton disregard of the treasure that is digital data. Most people just don't seem to care.
I think the problem is a bit more subtle. People do seem to care but feel powerless. In addition to that, complexity is high. When you tell people what is possible with massive data they label you as a conspiracy theorist because it's much easier to believe that Facebook is listening to you with a microphone than piecing that information together through thousands if individual pieces of data (especially when you're more likely to notice the hits rather than the misses). So now when they do record everything you say it feels like nothing changed.
It's exhaustion, it's impotence, it's desolate. But that's different than apathy. People care, they are just overwhelmed
Astute observation that "this" happens and is happening, regardless of whether you name it, like it, or participate.
So many people think splitting hairs about whether an LLM is a token predictor or an "agen will somehow change the impact they're having. Or people post about how they use, or will never use AI agents.
Surely the unwashed, tech-illiterate masses won't spam the internet with these things? Why yes, they can; And will. In the meantime, companies and early adopters, the crowd with the least ethics, will start ruining the show already.
You don't need a full air gap. Set up a microVM with network access limited to local network and send all package requests through a filtering gateway that only allows normal download endpoints. Or self host a big collection of popular packages if you need extra security.
No that's not what they did, they exposed jfrog raw. It would have been so extremely simple to gate services they need the llm to access... I mean, jfrog was not written with this kind of threat model in mind, and neither were a lot of other tools
Right, you mean it didn't go through a gateway? But would that actually have helped? The requests all went through jfrog didn't they? I guess it depends on the level of filtering at the gateway?
Whilst it might not be JFrog's threat model, I wouldn't assume it can be used as a full internet proxy.
I don't really mean to defend OpenAI here, but they did make some attempts at sandboxing. Although it does seem that they didn't really know what they were doing.
It would have been a case of isolating exactly what functionality is needed and wiring that up with the actual requests. Not like a full pass-through proxy. This is what we've been doing in our company as well
> jfrog was not written with this kind of threat model in mind, and neither were a lot of other tools
And we don't just magically know all the consequences of that.
Which is exactly why we do need full, physical air gapping. (Which, yes, would also include self-hosting a mirror of the package repo, if the point of the simulation is to see what's possible with the real package repo.)
> Agents need packages like the rest of us. Ruby gems, npm packages, Maven, pip, docker images..
Yes, yes they do, but read through artifact proxies are dodgy as fuck, which is why and facebook (and I assume a fuckload others) don't have them.
Also semi-airgapped labs are a lot less expensive than you think at that scale. Once you have to do multi-region VLANs with machine certs before you get access to juicy VLANs, the difference between "no internet for you" and "mostly airgapped" falls to almost zero.
Also I would want an artifact mirror because a) that give a good signal about how the model reacts, and what training material its latched onto, b) it hides what the models are doing from the outside.
Just let them communicate on the Bitcoin blockchain. "We" would have to freeze the chain and lose access to "our" billions of wealth, so not going to happen.
"Oh it's able to issue commands too fast." "Oh no, I give it full map access and it uses that."
Ego shooters are, naturally, also not good benchmarks.
OpenAI won DotA2 in 2019, a way better game.
reply