Tweet

This Thread may be Removed Anytime!

Twitter may remove this content at anytime! Save it as PDF for later use!

More from @perrymetzger

Perry E. Metzger

@perrymetzger

Feb 20

There are two common kinds of arguments critics make against Drexler’s molecular manufacturing. One of them is that it’s physically impossible, and I think at this point, that’s been disposed of pretty thoroughly. (At the very least, the existence of biology falsifies the claim.)

The other is that, although it might be physically possible in theory, it’s just too difficult to do in practice. However, so far, no goal that is within the bounds of physics seems to have been beyond our civilization’s grasp.

Indeed, if you learn enough about things like the manufacturing of the semiconductor chips used in the smart phone in your pocket, it sounds like pure science fiction, and obviously impossible, except, the thing is sitting in your hands.

Read 7 tweets

Perry E. Metzger

@perrymetzger

Feb 18

@ESYudkowsky

🧵Want to learn more about one of the two technologies @ESYudkowsky believes will soon end all life in the universe? Look no further than “Nanosystems”, one of the most interesting and important books I’ve ever read! a.co/d/j43PEZU

(Here’s an excerpt from the first page to whet your appetite.)

“Nanosystems” is a tour-de-force, an insanely thorough exploration of what you could do if you could build machines in which every atom is bound in a precisely defined location. (If you’re reading this, you yourself are, to a first approximation, an example of what’s possible.)

Read 11 tweets

Perry E. Metzger

@perrymetzger

Feb 17

@alyssamvance

@alyssamvance @ESYudkowsky Unfortunately, the people who claim to have been worried all this time have (de facto) done nothing, spending what effort they did expend on efforts that were predictably ineffective, so what are you satirizing?

@alyssamvance

@alyssamvance @ESYudkowsky A bunch of people literally got enough money to work full-time on the problem without distractions over many years. This was a best case scenario! And yet, not much was accomplished. Whose fault is that?

@alyssamvance

@alyssamvance @ESYudkowsky And why should we believe if another 10, or 20, or 50, or 200 years was expended on this “activity” that we would get any substantive advances? Why would future performance by this team be any different from past performance?

Read 4 tweets

Perry E. Metzger

@perrymetzger

Dec 10, 2022

“AI will replace search” reminds me of “better agriculture will mean we can feed horses more cheaply than before, making horses cheaper”, except now we all drive cars. Only the lack of foresight seems even more extreme.

The disruption of Google’s business model is likely one of the most boring consequences, much as not being able to find your pen might be one consequence of a nuclear blast hitting your town. Sure, it’s probably a thing. Is it likely what people would see as most salient? No.

AI doesn’t just disrupt an industry, or even all industries, it disrupts *everything you know about being a person*. (It’s thrilling, scary, fascinating, and literally beyond our imaginations, all at once.)

Read 6 tweets

Perry E. Metzger

@perrymetzger

Dec 9, 2022

A couple of (badly organized and insufficiently well explained) AI risk comments, many of which are going to gel into a blog post in the next few weeks.

There are vastly more ways you could fall over or badly injure yourself when crossing a room than ways you won't, and yet you cross the room successfully.

That's because you're not randomly selecting among all possible ways to cross the room, you trained yourself very carefully in infancy only to pick ways with high probabilities of a good outcome. (And to be sure, sometimes people stumble, but not often.)

Read 21 tweets

Perry E. Metzger

@perrymetzger

Dec 7, 2022

@ESYudkowsky

I agree that just scaling GPT like systems doesn't produce AGI. I disagree with @ESYudkowsky (vehemently so) that we should all assume that the only reasonable outcome is we're all going to die.

https://twitter.com/esyudkowsky/status/1600345866866335744

And being direct, I disagree (again, vehemently) that delaying the development of this technology would have done some good, or that we can figure out how to deal with the risks without concrete examples of the things in front of us.

Very intelligent people are, for better or worse, the most likely to not only believe in their own totalizing ideas, but also to follow them to their conclusion no matter how bad the ideas might be.

Read 6 tweets

Share this page!

Perry E. Metzger

People who liked this thread also liked...

Try unrolling a thread yourself!

More from @perrymetzger

Perry E. Metzger

Perry E. Metzger

Perry E. Metzger

Perry E. Metzger

Perry E. Metzger

Perry E. Metzger

Did Thread Reader help you today?

Don't want to be a Premium member but still want to support us?

Send Email!