Forget Programming, AI Psychology Is Where It's At

In Memoriam

A Conversation with Stephen Wolfram, Part 2

We had a technical glitch on Friday and had to scratch our scheduled Twitter Space with Disco.xyz CEO Evin McMullen. We apologize for any disappointment.

The good news is she's agreed to come back next month to talk with us about what was intended to be the subject for today's newsletter: How to Prove You're Human in an AI World.

While many people believe crypto and AI are not closely entangled, we believe the opposite, that the flood of fake media enabled by generative tooling is going to urgently demand a secure and verifiable digital identity protocol to prove the video, audio or text you're seeing is real. Blue Checks it is not.

Since we didn't get around to the interview last week, we are improvising today, sharing some additional excerpts from our fascinating March 23 interview with computer scientist, physicist and mathematician Stephen Wolfram, in which you'll learn why programming is a dead end career, what new skills will be demand, why AI is lot like ancient Greek mythology, and why humans just aren't all that smart.

You can find the complete audio recording here, and previous excerpts from our Friday newsletter here.

Don't Go Into Programming

What should people learn to be ready for the future AI world?

One thing they probably shouldn't learn is programming, because programming, sort of low-level programming of the kind that's kind of tell the computer what to do, this step, this step, this step, that is in the process of getting automated.

Just as when I first started using computers, assembly language, everybody should learn assembly language. That's the only serious language you can learn. Very, very few people program an assembly language anymore.

Do Become an AI Psychologist

It's a strange thing and this question is sort of, What does it take to be an AI Wrangler or an AI Psychologist? You know, how do you get, how do you align your mind enough with the sort of AI mind to be able to say what you should put in the prompt?

As we were doing prompt engineering for this plugin for today, it's kind of a very weird experience because [when you] tell the language model to do this, you're telling it to do it in English. And it doesn't do it sometimes.

And then you put it in capital letters and then it does it a bit more. And then you phrase it differently and then it does it maybe a bit less. It's very strange.

In terms of this whole question about will humans have nothing to do anymore, this is one of the new job categories, prompt engineer, AI psychologist or wrangler. These are new activities for people that have just popped up in the last few months.

Learn How to Think Computationally

I think that part of it is to learn how to think computationally, because that's really useful.

And it's not what people learn when they learn computer science, aka programming, so to speak. It's something that actually is not yet terribly widely taught.

Getting to the point where people have a good grasp of how to think computationally about things, that's a really important thing to have because that's something if you're going to be a stakeholder in the computational, you know, contract that is the A.I. Constitution, you've got to be able to have sort of a computational way of thinking about things to be able to contribute to that.

Will AI Destroy Us?

The civilization of AIs is very similar to the experience we have with interacting with nature.

You know, nature does its thing. We sort of coexist with nature.

Occasionally nature throws up a hurricane or something and it causes us a bunch of trouble. We have at least managed to evolve to sort of coexist in a fairly stable way with nature,

AIs are coming at us a bit more quickly, and you know, exactly how that sort of co-existence will work is unclear to me.

I think that if there ends up being the one centralized AI, and that's all there is, that's a fairly unstable, brittle situation.

If it ends up being there are lots of AIs, and they form an ecosystem of AIs, that I think tends to be a much more stable situation.

On AI and Greek Methodology

One image that came to my mind recently about the way to think about this is like in ancient Greek mythology.

There were the Olympian gods, hanging out on Mount Olympus, so they were duking it out and having all kinds of dramas and so on. And occasionally the dramas that they had had an effect on the mortals, so to speak.

And perhaps that's a little bit of a model for what we can expect to see in the case of the ecosystem of the AIs.

On Why People Aren't That Smart

If the question is, will we remain kind of the computationally smartest things on the planet forever, the answer is we're sort of already not.

And it has nothing to do with AIs.

It's just a matter of there are lots of computations going on and lots of systems around us. Some of those computations are going on in our brains with the electrochemistry of brains and so on.

Some of them are going on in the patterns of fluids in the atmosphere producing weather, things like that.

It's a very human-centric point of view to say the kind of computations that happen to go on in our brains are somehow fundamentally more significant than the ones that are happening in nature.

And in fact, from a pure scientific point of view, we know pretty clearly now that there isn't really a way to say, oh, these computations that happen in brains are more sophisticated than the computations that happen in nature.

So we kind of already lost that particular game.

Create

Metaverse Fashion Week launches this week. March 28 to 31. Traditional brands like Adidas, DKNY and Dolce & Gabanna will mix it up with digital fashion companies like DressX and The Fabricant. Here's a rundown on what you can see and do. Don't miss a Wednesday panel on the future of digital fashion that we're moderating.

"Turn the Table" by German photographer Nana Hank is part of the "Diamonds in the Rough" curation on Foundation. No AI here. Real cameras, real light, real people.

"Howl" by Diela Maharanie, a visual artist based in Indonesia, is currently on auction at Foundation.

"Pushing Daisies" is an affordable open edition by French artist McBess whose work often hits 10 Eth or more.  

And finally, here's one of the many creepy outtakes made while creating today's cover.

Compress

Twitter source code leaked online, court filings show
Pieces of the computer code used to run Twitter were leaked online, according to court filings.

Apple’s Best Hope for New Headset: a Smartwatch-Like Trajectory
The company gathers top 100 executives for demos of mixed-reality device before June unveiling.

Microsoft: ChatGPT Shows Human-like Intelligence
On Wednesday, Microsoft researchers released a paper on the arXiv preprint server titled “Sparks of Artificial General Intelligence: Early experiments with GPT-4.” They declared that GPT-4 showed early signs of AGI, meaning that it has capabilities that are at or above human level.

Cool Tools

AI chatbots compared: Bard vs. Bing vs. ChatGPT
An AI chatbot chat-off.

Interrogate legal documents with ChatGPT
User claims he used GPT-4 api to 'chat' with a 56-page legal PDF document about the famous supreme court case: Morse v. Frederick

How did we do today? Love us or hate us, let us know.