Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Who manages the agents?

off-policy.com

Post reply on HN
Print view View original

101–108 of 108 posts

Re: Who manages the agents?

#101

Post by getrelaydesk »

dominotw wrote: post #96
jagged-chisel wrote: post #66

Earlier quoted context omitted.

I found a scam online about 18 months ago that was selling access to some "AI trading platform" that was supposed to have the kind of bot you mention. I used the live chat to talk to someone and asked about demonstrable proof - like given data up through the end of last month, make picks for this month, then look at this month's data to see how it would have done. "Then this product is not for you." Right, not for an…

but wouldnt that depend on your pompts though?

[flagged]

Re: Who manages the agents?

#102

Post by senordevnyc »

simonw wrote: post #100
senordevnyc wrote: post #97

Earlier quoted context omitted.

I don’t know, I can imagine that the equivalent of a PIP for an agent would be noticing flaws in its output and then creating additional evals, modifying the harness, upgrading the model, etc. to improve its performance. That’s not SO different from the PIP, except that the LLM doesn’t really care in the same way?

The person making the tweaks to the harness is the one taking accountability for fixing those mistakes - they're the DRI in this scenario.

Like the manager being responsible for someone on a PIP.

Obviously they’re different, but it’s interesting to think about what is unique about humans that makes them able to be accountable or responsible in a way that LLMs cannot be. Is it at core just that they can be fired? Or essentially the threat of suffering?

Re: Who manages the agents?

#103

Post by pizzafeelsright »

wwind123 wrote: post #47
blauditore wrote: post #11

The closer I start working with and on AI stuff, the more I start seeing the disconnect between doomsday predictions of what AI will replace vs. what it is actually capable of. Yes, it can do stuff, and yes, it's getting better. But the closer you look the more clear it becomes that the enthusiast vision of completely independent AI systems is unrealistic as of today. Yes, all the tech companies are pushing for exact…

I think a main worry is that, this AI wave is quite different from past technological revolutions in that, this wave is happening so fast, the speed that humans learn new skills and master new jobs would lag more and more behind the speed that machines replace humans in those jobs. Without societal or legal constraints, capital chasing the max efficiency and profit would just replace humans with machines whenever mac…

Dropping a non-native person into the midst of a superior technologically and culturally different environment puts them at a disadvantage. How do you legislate on-ramping someone who goes from living in 1926 to 2026? How does a company hire someone who cannot read the native language?

Re: Who manages the agents?

#104

Post by daveguy »

senordevnyc wrote: post #102
simonw wrote: post #100

Earlier quoted context omitted.

The person making the tweaks to the harness is the one taking accountability for fixing those mistakes - they're the DRI in this scenario.

Like the manager being responsible for someone on a PIP. Obviously they’re different, but it’s interesting to think about what is unique about humans that makes them able to be accountable or responsible in a way that LLMs cannot be. Is it at core just that they can be fired? Or essentially the threat of suffering?

I think just the possibility of consequences that they can give a damn about. An LLM doesn't have feelings no matter how much human-like text it can simulate. There is literally nothing going on between prompts for any given model. They are incapable of worry or any other emotion. Wipe the context clean and the LLM is completely unaware there was ever a problem.

Re: Who manages the agents?

#105

Post by narrator »

whattheheckheck wrote: post #99
narrator wrote: post #22

Earlier quoted context omitted.

An exec can ask an expert to explain themselves, and if they're a good CEO they can weigh the pros and cons of the arguments from their staff and make a decision. Someone that's much smarter or more skilled in a particular discipline can usually explain what's going on to someone that's not as smart as long as the gap isn't too big. However, I've found that with Fable, and some of the latest frontier models their cri…

https://arxiv.org/abs/2606.16475 Pretty sure it can dumb it down to cave man speak for any concept

Can it dumb down Ramanujen equations so someone with average intelligence could understand them if even world class mathematicians struggled to grasp them? This is the best analogy for super intelligence. Alpha Go can't really explain why it makes the go moves it does in a way a human can understand. Sure it could show them all the weights in the model, but that's not anything a human can understand.

Re: Who manages the agents?

#106

Post by sxiong »

blauditore wrote: post #11

The closer I start working with and on AI stuff, the more I start seeing the disconnect between doomsday predictions of what AI will replace vs. what it is actually capable of. Yes, it can do stuff, and yes, it's getting better. But the closer you look the more clear it becomes that the enthusiast vision of completely independent AI systems is unrealistic as of today. Yes, all the tech companies are pushing for exact…

I agree that's the status quo right now. AI does ~80%-90% of the job well but in many cases, users require 99%+ reliability and the work still requires a human. The real scare is that the projection of AI intelligence will reach ASI and then that gap will close or be marginally insignificant.

The vision of independent AI systems will eventually come to fruition, just not today. Not yet.

Re: Who manages the agents?

#108

Post by whattheheckheck »

narrator wrote: post #105
whattheheckheck wrote: post #99

Earlier quoted context omitted.

https://arxiv.org/abs/2606.16475 Pretty sure it can dumb it down to cave man speak for any concept

Can it dumb down Ramanujen equations so someone with average intelligence could understand them if even world class mathematicians struggled to grasp them? This is the best analogy for super intelligence. Alpha Go can't really explain why it makes the go moves it does in a way a human can understand. Sure it could show them all the weights in the model, but that's not anything a human can understand.

Yeah it can...
Post reply on HN