14 Comments

This is a brilliantly mind expanding post.

Expand full comment

Who do you know who has been training Chat-GPT3 on their own writings, and are they willing to talk about it?

Expand full comment

Here is one example of an author training the boy on their book. Code is open source - https://askmybook.com/

Expand full comment

Hi Brad, happy to discuss this. Please reach out and we can schedule a time: monica.amadio@stockton.edu

Expand full comment

This is something I am very interested in as well.

Expand full comment

Me as well.

Expand full comment

"there is also evidence that things have slowed down in the last few decades."

There is a lot of talk of Moore's law ending. Do you think it is possible that we have an incredible few years of AI development, but that Moore's law hitting a limit will begin to slow AI progress before we hit AGI?

In other words, is the current AI boom the last hurrah of the current (information processing) technological wave, and we are then waiting for the next technological wave (syntheic biology perhaps) to arrive?

Expand full comment

A good one! The abrupt ending suggest that it was ChatGPT which/who wrote this piece…. or this answer. Who knows?…. Who knows if Chat GPT has a sense of humour??

Expand full comment

I am guessing you are aware that some of the people behind Singularity are also with Longtermism thinking -- like MacAskill and Bostrom (https://www.vice.com/en/article/z34dm3/prominent-ai-philosopher-and-father-of-longtermism-sent-very-racist-email-to-a-90s-philosophy-listserv) and of course crypto wonder boy Bankman-Fried. Singularity and Longtermism often seem like a nice idea ... if you are a 15-year-old Ayn Rand fan.

One problematic aspect of "accelerating growth" and the impact of "AI" is how you measure such growth. Are you talking about an absolute value or relative growth? Bostrom came up with some bullshit formula like this that lacks any empirical or sociological insight.

Growth in intelligence = optimization power / recalcitrance

what on earth is this??

Expand full comment

That seems like an ill placed correspondence. Yes, you can find many of the people from the longtermism movement also talking about the singularity - but that's more a function of the fact that those people more often than not come from the tech sector, and that's pretty much *the* thing everyone is talking about. But the singularity is a point in time, not a movement. You are conflating the two and getting angry it seems.

And to be fair, I find the longtermism idea absolutely awful. Also, I hope and pray that AGI could be used to benefit ALL humans, truly, and not just the 1%. I have little faith it will though because many of those in control of the big systems now seem to have other motives - even as they spout platitudes. Watch for UBI. Until we see a real movement in higher circles to get that implemented, we will be in a growingly perilous position as a species.

Expand full comment

thanks for the thoughtful reply. yes, sorry for the outburst!.

Expand full comment

The ones that really need to play with AI more than anyone are our kids. It always seems with any new technology, the "natives" with that technology are completely next level users compared to people who learn new tech as adults. This was true with PC's, true with the internet, and very true with smartphones.

I noticed the other day that my youngest was asking the google how to spell words to put into a typed search field on his tablet, it occurred to me what a profound use of technology that was, figured out by a 5 year old.

It is now entirely possible for a child to "write" (via AI assist) the very first book they read themselves on their own. At very a fundamental level, everything is different.

Expand full comment

Not sure who researched "elevator operator" being only job no longer listed in Census occupation codes. My hunch, "typesetter" also vanished ... and a little digging so far, may confirm that. In any case, your article is insightful and a bit startling, of course. May have to subscribe.

Expand full comment