"It is difficult to get a man to understand something, when his salary depends on his not understanding it."

In this case, Dario Amodei cannot understand a world, where AI fails to deliver on his own promises. Amodei is a salesman, just as much as any CEO.

From where I'm standing, the very low wage underclass arrived many years ago.

AI and robotics just make it worse.

But it is very arrogant to think that it will be limited to certain types of jobs.

Things have never been meritocratic. We have always had extreme inequality. Technology has made things slightly more fair but that is still very unevenly distributed.

We _should_ be able to leverage advanced technology to lift everyone up.

I am going to point out something uncomfortable: I think that racism, classism, and elitism is extremely prevalent globally and may be one of the biggest impediments to the even distribution of technology benefits.

We do need to redesign society. That starts with having a realistic educated respect for human beings in general otherwise it's going to be a bad design. It also necessitates refined and contemporary worldviews that properly integrate technology rather than outdated vague ideologies.

> AI could create a permanent underclass of workers

AI can't create a permanent underclass, politics creates the permanent underclass by deciding to under-tax windfall profits and then under-resourcing a segment of the population.

The US seems culturally ill-equipped to deal with this reality. We have encouraged several generations of people to channel all of their talents into maximizing their individual income, regardless of externalities or impact to their community. There is low trust and minimal social reward for giving back. We idolize the loudest, most ignorant voices only because they are wealthy and famous. In my own work with the next generation of tech workers, this seems obvious. The younger generations see it as a zero sum game. You only win by making as much money as possible, and the ends justify the means.

> But Amodei argues that AI increasingly matches the full range of human cognitive abilities, so it will take away jobs drafting memos, reviewing contracts, and analyzing data that might otherwise emerge. A customer service rep who retrains as a paralegal would find AI waiting there, too. "AI isn't a substitute for specific human jobs but rather a general labor substitute for humans," he wrote.

A) This is a bunch of marketing hype disguised as performative commiseration. "Our product is so good we need to think about changing laws". Maybe that's true, but let's not get ahead of ourselves. We don't even know the true costs associated with AI or how much better it's even going to get.

B) All of these previous inventions were also touted as "general labor substitutes" and none of them ended up being true.

Let me make an alternate case - coffee in this country used to be an entirely automated process. Everyone has a dedicated coffee robot in their house. But our tastes have so shifted that now the average person gets multiple cups of handmade coffee a week. An entirely new job category called "barista" was introduced, and today we have over half a million of them. They are not high wage jobs, but they are comparable to something like the customer service rep job that Amodei is apparently worried about.

Even if AI were to take away massive swathes of white collar jobs (I'm still skeptical), the historical expectation is that new, unforseen labor categories open up. Just like nobody inventing a computer thought QA tester or video game streamer or spreadsheet technician was going to be a job in the future.

It's like an inverse Baumol's Cost disease - if AI does tank the value of all of these services, all of the services that require, I dunno - physical hands, go up in value. All of the niche things AI can't do suddenly become all the more valuable.

Let’s get a reality check. Current agent performances are no where near that kind of threat, and it’s uncertain when we get there.

I think people continue bullshitting in this domain because they’re worried they have no moat, so they have to discourage via sophistry or bold claims.

Don’t take my word for it, here are the stats https://scale.com/leaderboard

For reference, 25% means getting 1 in every 4 question right.

Secondly, if we ever do get to that point, people have lots of social mechanisms they’ve developed over millions of years of evolution which will kick into place. I am not sure anyone dreaming of this “utopia” is then going to be around to see it.

White collar will drop, blue collar will go up. Somebody will coin "AI quotient", a person's ability to make efficient use of AI tools in what they are doing.