Why should we expect everyone to be both a great researcher and communicator? The obvious result is that it is as effective as engineering managers. Sure, there's some amazing ones, but most aren't. Though that also doesn't mean no expertise in the field (i.e. non-engineering manager) is any better. It is just that managing/communicating is a different and orthogonal skill.
What needs to happen is we need to make it okay for people to specialize in more things. More nuance to this rather than trying to throw everyone into nice easy to manage buckets. Those buckets are just unrealistic abstractions filled with hope, denial, and laziness. Reality is surprisingly complex. Math can do a really good job helping you understand that, but it's a sufficient condition, not a necessary one
This feels backwards. I frequently joke "physics is the subset of mathematics that reflects the observable world". Math can be as abstract as it wants, but physics has a constraint. It must model an observable world (related, this us part of why people say String Theory is math and not physics)
Suppose you're a firework company and your fireworks blow up, burning down the entire town. Could the company be sued? What is considered reasonable safety measures?
IANAL, but I'm pretty confident there would be a lawsuit. Who gets charged might differ, depending if it is the firework factory that didn't take adequate safety precautions or a chemical supplier or someone else. If there wasn't an ability to sue that would be fucking crazy and we should all get up in arms about it. And isn't insurance supposed to be there to help mitigate the damages, regardless of fault?
Personally, given how it seems OAI's agents have been getting through either pretty obvious places (e.g. /etc/hosts) or that there wasn't close monitoring of the most obvious places (e.g. DNS, artifactory), I'd imagine it wouldn't be hard to find them negligent. Even if a single employee is to blame then are they not to blame for not monitoring the agents regardless? Unless the story is that the employee intentionally circumvented defenses (why?) then it seems it would be on OAI. But again, IANAL, I'm just someone who think if we can't sue we can sure riot until we can
There's a difference between being able to be successfully sued (civil liability, petitioned by a private entity) and charged (criminal liability, or petitioned by a government entity).
The thresholds for suing and charging differ greatly depending on the circumstances.
Another set of hypothetical examples that make things muddier:
- If I drive a fishing boat into a pier, I am liable, not the manufacturer of the boat
- If I drive a car over someone lying in the road, I am liable, not the manufacturer of the car
- If my life is in danger and I shoot a gun and kill my attacker, neither I nor the manufacturer are liable so long as I obeyed the relevant self defense laws and gun possession of whatever jurisdiction I am in
- If I fire a gun into a crowd indiscriminately, I am liable and several jurisdictions have used that to also hold gun manufacturer liable as well
That last example has been less successful as of late, but there are other variations too.
I love the idea of robot helpers but I don't love the idea of putting cameras all over my house and having ad companies surveil my intimate space.
I do not think people are even recognizing the security ramifications of these devices. A picture of your key is enough to replicate it[0], and yet we're just going to put these devices everywhere? LLMs have been pretty easy to jailbreak too, you think thieves aren't just going to get these robots to open the doors for them? Or what about governments? Even if you trust your government now, will you trust them with data where they can literally see everything you do and say? That's well beyond 1984. Or just what about what we see from top labs like OpenAI, Anthropic, or even Google and Apple? These trillion dollar companies are not showing themselves to be good at protecting data. Obviously some better than others, but this is a whole new level of intimacy with the data they'll get. Everyone is pissed at Flock, and you want to put mobile Flock in your house?
No thanks, it isn't worth trading all that for not having to load up the dish washer and cook my own meals.
I can't believe these discussions aren't more common.
I also don't like the idea of putting cameras all over my house and having ad companies surveil my intimate space, and I also don't like to add further things that require electrical power, and/or that make noise, etc. This is in addition to the security ramifications that you describe and that you don't describe, and other issues.
People also underestimate how dangerous those humanoid robots are. They are hilariously breaking every OSHA rule written by human blood.
Many contemporary humanoid robots do not even have an e-stop and this is why you see clips where people frantically try to stop an unstable robot by grabbing it. At the moment, it’s a solution searching for a problem and blood.
I don't think anyone is using "autocomplete" in the way your phone does it. But it is shorthand for a much more sophisticated version that is built in similar ideas. And autoregressive models certainly have that in their core structure. But people are lazy and neither want to say a lot when few words work nor will they read long comments, even if more accurate
I want to clarify, most people are using "autocomplete" specifically to differentiate from how we humans operate. Sure, there is an autocomplete aspect, but it's not the core nor anywhere near the full story.
If you over simplify everything then everything is the same. You have to have nuance. So don't just argue for the sake of arguing. Take that same passion to deeply understand these systems. Take that same passion to try to understand yourself and those around you
I think people have more passion to argue than they have passion to learn. Probably doesn't help that SV culture tells people to hustle so hard that they don't have time to think. Gotta go fast?
You're talking past the parent's claim. If your axioms are wrong then of course your logic will be wrong.
But then again, maybe to support your point people frequently say "start from first principles" when those are usually the thing that needs to be found, not the place you start. But LLMs, like humans, love to be confident about things they aren't sufficiently trained on
> If your axioms are wrong then of course your logic will be wrong.
Equally, even if your axioms are right your logical execution can still fail.
Neither individual humans nor LLMs get either of those steps right as a default. Just as LLMs must use hacky CoT and other systems of error checking and correcting to even begin to emulate reasoning, humans must do the same thing.
We are apes. We are driven by emotion and solve problems either when we are forced to or as a sport. We did create the ultimate automata of logical execution: the deterministic von Neumann Machine, but a human brain in no way reflects that operation. Nor, as you have observed, does an LLM despite using one as its substrate.
Why should we expect everyone to be both a great researcher and communicator? The obvious result is that it is as effective as engineering managers. Sure, there's some amazing ones, but most aren't. Though that also doesn't mean no expertise in the field (i.e. non-engineering manager) is any better. It is just that managing/communicating is a different and orthogonal skill.
What needs to happen is we need to make it okay for people to specialize in more things. More nuance to this rather than trying to throw everyone into nice easy to manage buckets. Those buckets are just unrealistic abstractions filled with hope, denial, and laziness. Reality is surprisingly complex. Math can do a really good job helping you understand that, but it's a sufficient condition, not a necessary one
reply