AI agent sales has been a massive failure. Simple issues like its inability to distinguish a low quality lead from a completely wrong lead aren’t possible yet.
This isn’t agi. Or anything in the way to bring it.
virtually all the resources from oil, food and land, IP and tech (semiconductors), even human capital, and advanced IT. everything is captured already. from free laisure entertainment minutes, to internet search, to social. every single resource is captured and you are stepping on somebody toes. worse, most industries are monopolies/or-close, meaning couple whales dominate everything, and nobody else really matters.
whatever "new" pie comes out, it is usually at expense of something else.
this "creation of pie" is such an illusion. go and try to "create a pie". it is such an illusion.
just go and try to even grow food out of earth with sunlight and water (which all should be free), yet farmers notoriously unprofitable and would not survive without government subsidies.
Ground truth is not consensus, it has to be graded against what actually works for the original goal. Plenty of scenarios with AI and Humans can result in consensus around incorrectness.
While pedantically correct, I think the comment above assumed that you've correctly specified the work. If you can't correctly specify your work, AI agents are just going to help you get a non-solution faster.
Isn't coding the act of specificying the work to a processor? And AI agents are supposed to bridge the gap with intelligence from less specificed to more specified or possibly even more intelligent and alternate implementations?
What I meant by "ground truth" is that it is not fuzzy, not AI-evaluated, and not a consensus. The test suite passes or it doesn't. The codebase lints or it doesn't. The performance improved or it didn't.
An agent can help you create the specification, but it's up to you to know whether it's correctly testing that you got the result you wanted.
This is really dissapointing release for such a promising technique. Long walks with fanned vectors can actually be token optimizing vs token burning when combined with self grading each agent along the walk and compared to manual long coding walks to solve first pass problems. But instead this frames it (assumptively) as a tokenmaxxing strategy. There are also many other strartegies that can prove effeciency and wider solution consideration with consensus, but none of this is explained why its an improvement or better than other technqiues.
Its like you guys aren't even aware of the primary problem you are all facing: your token burns aren't paying off anyore against standard coding -- and looking net negative. I have to ask, are you this unaware of your core problem set here?
There are no any examples, proofs, or scenarios that show why there is improvement either in complexity or reliability of the solution or effeciency to the path of the solution. I'm baffled.
It’s a stupid statement by a stupid philosopher. Years later we learned collective development and incentive produced a society he could have never imagined.
It’s simply being looted now by the idiots this moron worshiped.
It’s from the Melian Dialogue (https://en.wikipedia.org/wiki/Siege_of_Melos) and it’s perhaps the most succinct embodiment of the realist school of international relations/politics.
I quote it here because the best way to get a day off is not to continue being weak, but to find strength, just as miners and railroad workers found strength in the 1870s.
The commentary was that of a general/philosopher/historian fused together... and really just a statement of "the weak are meat". Yes power has an advantage but its not the end all be all configuration.
Maybe he meant this somewhat disparagingly, but ultimately not enough. For a greek.
But I would say that processor like the Z80 are so simple and well-known and documented in the wild that they would have some chance to be bootstrapped again with what would realistically survive, one way or another, in a doomsday scenario. But yeah 20 year is too short, it's more like 40 years.
Claude code was one person's idea as a pet project and now it's singlehandedly 5x'd Anthropic's valuation. Sometimes single people matter, that's life.
I think both are true. Claude Code was one man's experimental project, but it's an application of AI, not AI itself.
Anthropic is a large company, with thousands of employees, and seems to be 100% (maybe 200%) LLM and scale pilled. All the advances from one model generation to the next are the result of dozens of experiments first at small scale then at larger scale, all competing for the same "development compute" portion of their overall "development + inference" compute resource.
In this environment, even if there are researchers who have ideas not on the "LLM + scale is all you need" path that Dario seems hell bent on, there seems to be not much chance that these ideas can compete for resources and compute with the mainstream experiments that the company believes their future depends on.
Maybe an individual developer like Sutskever, engaged purely in research rather than manning a barely turnable oil tanker, can make a difference, but at a company like Anthropic it seems way less likely. Cherny's baby is 100% aligned with Anthropic's mission of selling LLM tokens. Someone else fighting the mission, trying to pivot Anthropic in a new better direction is not likely to have such luck.
Academia is fundamentally in for a long and unstoppable decline due to population changes and birth rates.
But I had assumed we’d end up with a bunching effect that would push up demand for MIT rather than down. (When there is an over decline in something, often remaining participants bunch harder into the most desirable remaining)
This isn’t agi. Or anything in the way to bring it.
We are in mass delusional state.
reply