comfysocks
comfysocks on Hacker News
Maybe he and Trump can flee to Russia?
If those people want to build a bubble for themselves to live in, then have at it. BUT DON’T FORCE YOUR BUBBLE ON THE REST OF US.
But frankly I’m calling bullshit on the “sensitive ears” explanation. There seems to be a focus on flagging articles that have discussions critical of the tech-billionaire/MAGA nexus. Also some of the loudest voices rallying to the defense of said nexus are accounts only a few months old with an unusually large amount of karma for the age of the account.
Here’s my theory about AVP.
Apple’s strategic position is to be the main device by which apps and AI can gain access to end users. This puts them in a nice position to charge a toll on everything that passes through their device. Now if someone were to come out with a new type of device and a new paradigm that could disrupt iPhone as the main device, Apple would be in trouble.
Not long ago, Zuck seemed really confident about metaverse/AR/VR. Apple took the disruption threat seriously and countered with AVP. Now it looks like metaverse flopped. I suspect the priority of AVP has gone down inside Apple accordingly.
Yes, “consistent with”. Thanks.
To be fair to Ed, I’d describe his usual argument (at least currently) as saying that Meta, MS, google are “mature” companies trying to be maintain the high valuations and growth of a young company, which they no longer are.
If you take this to be his argument, then dan’s numbers are more consistent ed’s claim.
Agreed. With steak, the main point is to let the quality of the meat speak for itself. Too much input from the cook and you just get in the way of this.
Absolutely. I don’t know how strong their case is or isn’t, BUT if you hire Jony Ive and a bunch of Apple device people in order to build something that is a threat to the “Apple tax” revenue machine, they are coming after you without a doubt.
It seems like the widely-covered news stories that support OpenAI’s narratives originate from things that happened inside OpenAI.
It wasn’t an outside benchmark evaluation, or an external security researcher that uncovered the rogue agent behavior at this time, it was OAI itself. It wasn’t a notable outside mathematician that used AI to disproved Erdos’ unit distance conjecture, but OAI itself.
Im not saying these things are fabricated, but maybe the curated result of an effort to shape a narrative.
And you need a rich person to do the exploiting. The power differential is a key ingredient.
> Can someone breakdown to me how this makes any sort of economical sense?
Capital markets are excited by AI.
By tying his rockets to AI with his vision of “orbital data centers”, Elon turned an $8 per share IPO (at least according to financial times and morgan stanley) into a $135 per share (1.8T) IPO.
Fair enough. I too use LLM coding tools all day everyday. If we’re specifically talking about the arguments in this thread, I have experienced both sides of the argument. Some use cases are more in-distribution than others. I wouldn’t dismiss either side out of hand. I see it more as a “yes, but…”
Somewhere, someone may be painting an equally reductive picture of you and your position.
You two are obviously smart people, but have you considered that maybe you are unconsciously being backed into a position that is just as absolutist as those you are arguing against? That you are just taking the opposite side in a binary argument where each side sees only one side of the coin?
There are many possible futures, but the one that will come to pass is the one we collectively allow to happen. To a degree, of course.
A future that I like to imagine is one where LLM scaling laws hit a plateau. Plateau-level LLMs become commodities. They get distilled into small sizes. Everybody can have plateau-level local LLMs. People discover that it’s better to be a local-powered independent centaur than to be a reverse centaur for a mega-corp.
I had a similar experience at a bank some time ago. To sign up, you had to sign a digital pad without seeing what you were signing first. You could get a copy mailed to you later. At that future time, I was told, you could you cancel the agreement if you found it objectionable.
Being a bank, this has nothing to do with HIPAA. Just a dark pattern.
A bullied kid finds refuge in sci-fi and fantasy books. This kid builds a mental fantasy world where they get revenge on their tormentors. In this mental fantasy world, every self-serving thing they do is “righteous”, because it undoes the harm that was done to them. Their manifesto is a mish-mash of ideas from the books, but twisted to make them into the good guy.
Some of these kids eventually grow up and meet people who are kind to them. They find positive lessons in real human social interactions. They leave their protective fantasy bubble behind. They eventually learn to seek justice where there was injustice.
Others never grow up. They end up seeking to fight injustice with a new form of injustice. Only this time, they get to be the tormentor.
Tesla has a problem. Its sky high valuation is based on the potential upside from “visionary” scifi-future programs. However, the narrative is losing credibility.
Once upon a time, Robotaxi and Optimus sounded like visionary futuristic ideas. However:
We now live in a world where Waymo exists, and autonomous taxi hasn’t proven to be as socially transformative as promised. No huge latent market exposed. Aslo, Tesla is lagging behind.
We now live in the world where Unitree exists, and is shipping! Again, tesla is on the back foot. The technology no longer seems quite so futuristic. Disruptors usually disrupt from down-market— that puts BYD cars and Unitree robots in the better position.
The new sci-fi furniture proposition is colonies on Mars. That one is likely to stay in the future for a long time. You think the Chinese will beat him to it?
If you apply Occam’s razor, this explanation stands out.
> "We need an approach to make sure AI doesn't destroy the world and wipe humanity to extinction."
"Yeah, and quotas on web scrapers!"
——
I see it as:
“Let’s do a study on abstract future concerns.”
Vs.
“Let’s take action on concrete present-day concerns.”
I hate this kind of thing too, but realistically this is every newspaper/digital that I’ve subscribed to ever. In a way, I get it. The new media landscape has put traditional news organizations in an existential crisis. I would hate to see world where the only way they could exist is under someone like bezos.
An easy way out might be to just ask the candidate how they got from one step to another. That way you can differentiate between “toxic, so smart i can’t communicate” and “non-toxic, so smart that i honestly thought the step seemed trivial, but would be happy to explain if i was mistaken about that”.
When the author says “crimes”, I didn’t take it to mean a literal criminal enterprise, but instead casual-speak for something like “bad governance”. It’s like if you saw some bad programming practices and casually called them “programming crimes” to avoid.
Similarly, when the author says “trash bag”, I don’t take it to mean a literal trash bag.
Sure, indiscriminate tokenmaxxing is a gamble that can pay off sometimes. However, I think that the decision to take any gamble should be made by someone who will bear responsibility for the downside as well as the upside. I would prefer to search for new usages in a more strategic way. I agree that experimentation is a great way to learn if done intelligently and with limits. Full “Monte Carlo” makes sense when ops are cheap enough. It seems some orgs don’t think tokens are cheap enough yet.
That reads a bit like: an equal role in external politics, but not in internal church politics. It’s hard to have a role in politics if you don’t have a voice.
>>I have found that, for many of the statements about what AI should do, I would actally be happier if the letters "AI" were replaced with "companies"
Perhaps, but at the moment AI is at the forefront of the pre-regulation land grab.
To me, the problem is when IP laws are stretched and abused by big corporations (like, say Disney) or by patent trolls. If IP laws could work in a way that gave limited and reasonable protections to actual creators and innovators, then I don’t have a problem with them.
But this new tool is not a blacksmith’s tool in the traditional sense. It’s more like an automated blacksmith that works fast, for cheap, does mediocre work, but has this mediocre skill level in an exceptional broad range of tasks.
Yes, and potentially extracts the wealth at the cost of the new grad’s job prospects.
Can you imagine a few decades earlier some former corporate executive giving a commencement speech at a US college extolling the virtues of offshoring, and how it will make his mega corp a lot of money?!
> I seriously doubt it. Scaling is already strained (don't buy into the "exponential" hype). And, in any case, the competition will be against the frontier models that will exist in two years.
But even if scaling plateaus for the frontier models, maybe distillation will improve to the point where smaller more manageable models can reach the same plateau. That would be great for local.