I'm getting increasingly worried and frustrated about this new phase of AI disillusionment.
It's naturally fine and healthy to notice the danger of hundreds of billions moving around in financing shell games, and the cost of inference being a wildcard. And to be concerned about that.
But I think too many people are conflating that with evidence that this whole thing has been another Crypto/NFT situation. So it's safe to ignore and make fun of.
I seriously worry about this because it separates people roughly into two groups:
1. The group who thinks AI is gross and scammy, and/or a bubble, and therefore they avoid it as much as possible
2. The group who sees it as a new, essential tool for upgrading your life and business, and therefore something that must be continuously used and integrated throughout your life
We talk a lot about separation of different groups in the U.S., or globally, and I think this is leading to the biggest possible separation.
People who are AI-native, or AI-integrated, and people who aren't.
This reminds me of reading. Talking to someone who reads 20-50 good books a year is completely different than talking to someone who hasn't read a book since they were forced to in school.
They see the world in completely different ways. And they usually have completely different preferences, experiences, and opportunities. Like they're from different planets.
Now imagine that with an exponent. Being AI-native is like the opportunities magnifier of being a voracious reader, but magnified many times over. Also doesn't help that they (anecdotally and not surprisingly) tend to be the same people.
I'm seeing this difference very acutely in meet people who are wired up vs. not. It's the vast difference in what they think is possible, and what they're shipping (doing) every week. How much they're building, or creating.
And I'm not talking about just tech or apps. I'm talking completely non-technical people as well. Even those people who are fully wired up with AI are just way more productive at whatever it is that they do.
It feels gross to me. It feels unfair.
And it makes me want to evangelize even louder for AI adoption. To try to make sure fewer people fall for the "it's a phase" narrative, or "it's all evil and bad for society" narrative.
I honestly think those people are screwed. Or at least seriously hobbled.
I guess my takeaway here is that we should, in my opinion, be thinking about this as a binary type situation. It's not quite that, but I think it's a useful frame.
You either get yourself and your loved ones into the AI-native camp, or you and they are likely to struggle massively in the coming years.
That doesn't mean thinking AI is Jesus, or that it's perfect, or that it's without moral complications, or dangers of over-use. You don't have to be a zealot about it.
But think of it like a pill that makes you smarter. Or a team of 100 super-intelligent interns who can help you do whatever you want.
Use them. And encourage the people you care about to do the same.
My guess is that there are lots of experts being asked about this from the industry side and also the government side.
Do you expect them to make public announcements of all that, all the time, as it's happening in realtime.
We're hearing that the labs are being called to talk to the government and we're assuming the absolute worst.
If the labs were just releasing these models, and companies were getting hacked, and infra was falling over, all these people complaining about the bans would be like,
"WHERE IS THE GODDAMN GOVERNMENT? WHY DIDN"T THE LABS GO TO THEM FIRST?"
They're vapid noisemakers.
Here's a more logical and less emotional way to look at what's happening with the chaotic regulation of these AI models...
It took decades/centuries to come up with policy around restricting weapons, drugs, heavy machinery, the purchase of chemicals, what's legal or illegal to search in a search engine, etc.
AI is now a new thing that we're going to have to regulate.
We'll need industry people to be part of that, and most importantly the government...since they represent us.
As stupid as any given push or pull of policy is in the next weeks and months and years, realize that this normally takes years or decades to get right.
You're watching policy take shape in realtime, for a tech that very few people even understand.
It's going to be nasty. Expect it.
And stop assuming the worst, of everyone, all the time.
@syndrowm This is all brand-new stuff happening in real time. If it takes decades for seatbelts and building codes and everything to be secured, do you really expect perfect policy to happen in the matter of days or weeks when the technology is developing by the minute?
Do you not get that we have to figure out how to secure this stuff we’re making?
Did you know it’s illegal to make and sell certain kinds of weapons?
You have to have licenses for using heavy machinery.
You can’t just make poisons.
The government massively restricts and controls all sorts of stuff.
This is just confusing because it’s brand new tech that just went from being safe to potentially dangerous.
It’s completely new ground for everyone.
And you’re screaming like a toddler playing with a new toy when an adult takes it away.
We are all the adults. You and me and everyone. We are the industry people, and the government and the group of experts that come up with the policies that we live by.
This Is absolutely brand-new technology. That’s why it is chaotic.
I'm developing a deep hatred for anyone who restrict our AI usage
Imagine if someone decided:
> whether you can drive a car
> if you can use the internet
> how much electricity you can consume
> IF you can own a smartphone
NOBODY WOULD TOLERATE THESE.
yet... for some reason... we let the U.S. government block us from using AI models?!
enough.
Joining Anthropic is like selling your soul.
They took all of Humanity's knowledge and distilled it into their own, closed models.
Now they use these models to generate insane profits for themselves
And they don't even release them to the public, it's just them, a handful of Big Tech customers, and the U.S. government
Which, if you knew the space deeply, you might realize is a very good thing.
The problem is not that somebody is trying to decide that level of intelligence for us. The problem is making sure that that group of people is the right group of people.
Currently, it’s a chaotic mix of the main industry heads, who are competitors, combined with some part of the government.
It is not well formed yet, but this is the type of consortium that we should ideally pursue for this.
You realize there is a group of people deciding what is safe for you to eat, right? It’s the same concept here. It’s just in early days.
If we strip away the headlines about Mythos 5, the story is brutally simple: someone, somewhere, is deciding what level of intelligence you and your company are allowed to access.
Prompts to Run When Fable Comes Back
My collection of prompts that benefit from the highest possible intelligence, and that improve your systems at the root level.
https://t.co/iVtiNJ8TnH
I'm getting increasingly worried and frustrated about this new phase of AI disillusionment.
It's naturally fine and healthy to notice the danger of hundreds of billions moving around in financing shell games, and the cost of inference being a wildcard. And to be concerned about that.
But I think too many people are conflating that with evidence that this whole thing has been another Crypto/NFT situation. So it's safe to ignore and make fun of.
I seriously worry about this because it separates people roughly into two groups:
1. The group who thinks AI is gross and scammy, and/or a bubble, and therefore they avoid it as much as possible
2. The group who sees it as a new, essential tool for upgrading your life and business, and therefore something that must be continuously used and integrated throughout your life
We talk a lot about separation of different groups in the U.S., or globally, and I think this is leading to the biggest possible separation.
People who are AI-native, or AI-integrated, and people who aren't.
This reminds me of reading. Talking to someone who reads 20-50 good books a year is completely different than talking to someone who hasn't read a book since they were forced to in school.
They see the world in completely different ways. And they usually have completely different preferences, experiences, and opportunities. Like they're from different planets.
Now imagine that with an exponent. Being AI-native is like the opportunities magnifier of being a voracious reader, but magnified many times over. Also doesn't help that they (anecdotally and not surprisingly) tend to be the same people.
I'm seeing this difference very acutely in meet people who are wired up vs. not. It's the vast difference in what they think is possible, and what they're shipping (doing) every week. How much they're building, or creating.
And I'm not talking about just tech or apps. I'm talking completely non-technical people as well. Even those people who are fully wired up with AI are just way more productive at whatever it is that they do.
It feels gross to me. It feels unfair.
And it makes me want to evangelize even louder for AI adoption. To try to make sure fewer people fall for the "it's a phase" narrative, or "it's all evil and bad for society" narrative.
I honestly think those people are screwed. Or at least seriously hobbled.
I guess my takeaway here is that we should, in my opinion, be thinking about this as a binary type situation. It's not quite that, but I think it's a useful frame.
You either get yourself and your loved ones into the AI-native camp, or you and they are likely to struggle massively in the coming years.
That doesn't mean thinking AI is Jesus, or that it's perfect, or that it's without moral complications, or dangers of over-use. You don't have to be a zealot about it.
But think of it like a pill that makes you smarter. Or a team of 100 super-intelligent interns who can help you do whatever you want.
Use them. And encourage the people you care about to do the same.
Remember when everyone said “businesses could never use AI because of how much they hallucinate.”?
Isn’t it weird how that’s mostly not an issue anymore?
Be very skeptical of people telling you what AI I will never do.