meta_pixel
Tapesearch Logo
Log in
The John Batchelor Show

2/2: #ArtificialIntelligence: Foundational Models argued to be genuine and perhaps unlimited competition for human work. 2/2: #AI: Birth of the Foundational Models. Ludwig Siegele, Economist. (Originally posted June 28, 2022)

The John Batchelor Show

John Batchelor

Books, News, Society & Culture, Arts

4.52.8K Ratings

🗓️ 5 January 2023

⏱️ 11 minutes

🧾️ Download transcript

Summary

Photo: No known restrictions on publication.
forbidden planet
movie
machinery
machine
technics
sweden
tekniska museet

@Batchelorshow


2/2: #ArtificialIntelligence: Foundational Models argued to be genuine and perhaps unlimited competition for human work. 2/2: #AI: Birth of the Foundational Models. Ludwig Siegele, Economist.
(Originally posted June 28, 2022)

https://www.economist.com/interactive/briefing/2022/06/11/huge-foundation-models-are-turbo-charging-ai-progress
Update Description




Erik Brynjolfsson, an economist at Stanford, worries that an obsession with scale and person-like abilities will push societies into what he calls a “Turing trap”. He argues in a recent essay that this focus lends itself to the automation of human activities using brute computational force when alternative approaches could focus on augmenting what people do. And as more people lose their jobs their ability to bargain for a fair share of the benefits of automation will be stymied, leaving wealth and power in fewer and fewer hands. “With that concentration comes the peril of being trapped in an equilibrium in which those without power have no way to improve their outcomes,” he writes

Transcript

Click on a timestamp to play from that location

0:00.0

This episode is brought to you by Slack. With Slack, you can bring all your people and

0:05.9

tools together in one place. It's your digital HQ where you can increase productivity,

0:11.1

enable flexibility and automate workflows. Plus, Slack is full of game-changing features

0:16.8

like huddles for quick check-ins or Slack Connect, which helps you connect with partners

0:21.0

inside and outside of your company. Slack, where the future works. Get started at

0:27.0

Slack.com slash DHQ.

0:35.0

This is CBSI in the world. I'm John Batchett with Ludwig Siegler, the European Business

0:40.5

Editor, former US Technology Editor at the Economist magazine writing about a new thought

0:47.4

of what a computer can do. This is artificial intelligence, artificial emphasis. However,

0:54.6

artificial is okay when you're writing code. I need to understand a GPT-3 with a code

1:00.1

X program and a co-pilot program. Does this mean that the code writing that is done for

1:07.0

I marvel at the video games that are getting more and more detailed? Is that what a foundational

1:12.6

model can do to write code for those creators?

1:16.6

That's one of the things it can do. When GPT-3 was developed or trained, people expect

1:24.4

to do all things, having to do with language writing, sentiment analysis and all that stuff.

1:31.2

Then they discovered that this thing can also code. If I tell it, write me or program

1:37.1

me this website, it would spit out the necessary code for that. OpenAI, which created GPT-3

1:46.9

and Microsoft, which is a partner of OpenAI, they developed a retrained and more and

1:54.3

it turns out it can now really program well. Co-pilot and codex, which you can do, you can

1:59.9

sit down and with your voice, you can tell the computer, okay, program this webpage or

2:04.5

this game or whatever. It does it within seconds and these programs are actually pretty good.

2:10.7

What Microsoft told me is that the people who use co-pilot, at least in the most important

...

Please login to see the full transcript.

Disclaimer: The podcast and artwork embedded on this page are from John Batchelor, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of John Batchelor and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.

Copyright © Tapesearch 2026.