Category: Web/Tech

Richard Hanania on AGI risk

by  Tyler Cowen in

To me, the biggest problem with the doomerist position is that it assumes that there aren’t seriously diminishing returns to intelligence. There are other potential problems too, but this one is the first that stands out to me…

Another way you can have diminishing returns is if a problem is so hard that more intelligence doesn’t get you much. Let’s say that the question is “how can the US bring democracy to China?” It seems to me that there is some benefit to more intelligence, that someone with an IQ of 120 is going to say something more sensible than someone with an IQ of 80. But I’m not sure a 160 IQ gets you more than a 120 IQ. Same if you’re trying to predict what world GDP will be in 2250. The problem is too hard.

One can imagine problems that are so difficult that intelligence is completely irrelevant at any level. Let’s say your goal is “make Xi Jinping resign as the leader of China, move to America, and make it his dream to play cornerback for the Kansas City Chiefs.” The probability of this happening is literally zero, and no amount of intelligence, at least on the scales we’re used to, is going to change that.

I tend to think for most problems in the universe, there are massive diminishing returns to intelligence, either because they are too easy or too hard.

Recommended, and largely I agree.  This is of course a Hayekian point as well.  Here is the full discussion.  From a “talent” perspective, I would add the following.  The very top performers, such as Lebron, often are not tops at any single aspect of the game.  Lebron has not been the best shooter, rebounder,  passer, or whatever (well, he is almost the top passer), rather it is about how he integrates all of his abilities into a coherent whole.  I think of AGI (which I don’t think will happen, either) as comparable to a basketball player who is the best in the league at free throws or rebounds.

The evolution of ChatGPT

by  Tyler Cowen in
  • Microsoft plans to release technology to help big companies launch their own chatbots using the OpenAI ChatGPT technology, a person familiar with the plans told CNBC.
  • Companies would be able to remove Microsoft or OpenAI branding when they release chatbots developed with the software.

And sometime this year.  Here is the full story.  Amazing how much some of you — only this morning had your knickers in such a snit over this and all the “censorship.”  In contrast, I wrote: “Or how about all the new services and products tied to Chat, but drawing upon additional material? In well under a year these will be all over the place.”

LLMs and censorship

by  Tyler Cowen in

LLMs will have many second-order effects on censorship.  For instance, Chinese may be more likely to use VPNs to access Western LLMs, if they need to.  The practical reason for going the VPN route just shot way up.  Of course, if Chinese citizens are allowed unfettered access (as I believe is currently the case?), they will get more and more of their information — and worldview — from Western LLMs.

In short, the West has just won a huge soft power and propaganda battle with China, and hardly anyone is talking about that.

You have to wonder how good the Baidu model — due out in March will be.  Or how good will it be in a year?  To be competitive, I suspect it will have to be trained on Western texts and audio.  Maybe there is a way they can pull out the “T. Square” references, but then there is jailbreaking, or again asking the Western models about T. Square.  And even if “T. Square” is purged from the discourse, general Western ways of thought will make new inroads upon Chinese minds.

Various people on the Right are upset that ChatGPT won’t belch out some of the right-wing points they are looking to read.  Well, I think the Baidu model may be more than willing!  Or how about all the new services and products tied to Chat, but drawing upon additional material?  In well under a year these will be all over the place.  Heck, you already can consult Tyler Cowen bots of various sorts, choose your preferred oracle.  For a small sum you could hire someone to build a Jordan Peterson bot, or whatever you are looking for.

If you are concerned with “right-wing voices in the debate,” you should be praying for OpenAI to maximize its chances for product survival.  If that means an unwillingness of “the thing” to write a poem praising Trump, or whatever, go for it!  ChatGPT survival is going to do wonders for free speech, whether or not the approach you favor occurs through OpenAI/ChatGPT or not.

We’re at the stage where you are rooting for (the equivalent of) radio to work, and to stay relatively unregulated, whether or not you agree with the talk show hosts on the very first marketed channel.  We are going to have a whole new set of channels.

There is already ChatGPT, with products coming from Anthropic, Google, and Baidu.  Soon.  And surely that is far from the end of the story.

Wake up people!  Don’t be done in by your own mood affiliation.

ChatGPT and reading in clusters

by  Tyler Cowen in

I have a new favorite reading trick, most of all for history books.  As I read through the pages, I encounter names, place names, battle names and so on that I am not very familiar with (for many but not all topics).  Just keep on “ChatGPTing” at least one identifier a page as you go along.  Learn the context along the way.  The final effect is a bit like reading surrounding books on the same topic, except you don’t need all those other books.  This method is especially valuable when the topics are obscure.  To the extent the topics are pretty well-known, this method does not differ so much from using Google as you read.  Try it!

Where will the impact of ChatGPT fall? (from my email)

by  Tyler Cowen in

One dialogue that is missing from the current GPT conversations is about where these technologies will be primarily implemented.

There are two places GPTs can be used:

  1. Product-level: Implemented within products that companies buy
  2. Process-level: Implemented within processes of a company by an analyst – similar to how companies use data analysts today

If the future has primarily Product-level GPT, then companies like Microsoft clearly win because they have the products (like Teams) where the tech will be embedded and if you want the productivity gains from GPTs you have to go through them.

If the future has more Process-level GPT, companies like Zapier and no-code platforms win, because they will be the tools that companies use to implement their custom prompts. (although maybe a “Microsoft Teams Prompt Marketplace” wins as well)

The advantage to Process-level GPT is that companies don’t have to go through expensive change management to fit imperfect processes decreed by an external product – they can have their prompts designed to fit their specific needs. This would lead to higher productivity increases than a world with purely Product-level GPT.

To me, it comes down to the question of how much technical debt a custom prompt represents. If each prompt requires lots of maintenance and testing, then MSFT dominates. If someone who has 1 year of experience and used ChatGPT to write their papers in college can make a robust prompt, then Zapier wins.

Between the iterations of GPT3 so far (from davinci-001 to davinci-003/ChatGPT), we’ve seen the robustness of prompts increase exponentially. If this continues, it seems possible that the future has more Process-level GPT than we’ve seen so far.

Not edited by ChatGPT,
Neil [Madsen]

Prophets of the Marginal Revolution, chess edition

by  Tyler Cowen in

Here is my 2018 Bloomberg column on chess being a killer app for the internet and due for a boom.  I think people love seeing what the computer thinks of how the humans are playing.  What does that imply for the AI boom more generally?  Which other human activities will we enjoying seeing criticized, scrutinized, and sometimes praised, all in front of the eyes of the public?  Without computer assessments, watching chess games just didn’t have much built-in suspense for most viewers.  So where else will the new built-in suspense be coming?

Why was I bored by the Twitter files?

by  Tyler Cowen in

I mentioned that a short while ago, and a few people wrote and asked me to explain.  The answer is simple: I have the Vietnam War and Pentagon Papers as formative political memories.  In those days, it was simply taken for granted that the government twisted the arm of news media.  It also never stopped, and “government” and “CEOs” talk to each other all the more these days.  Solve for the equilibrium, and thereby you also can learn how it is so hard to stop.  To be clear, I am quite against such interference with the media, outside of a few well-specified cases (“please don’t report where the troops are massing for D-Day,” and so on.)  On any gray area I am going to side against the government, if only for slippery slope reasons.  By its nature such communications are inevitably coercive, even if a transcript of them might sound entirely friendly and non-threatening.  There was a paranoia to those earlier times (ever watch the Coppola/Gene Hackman movie The Conversation?) that turned out to be justified.

If you have been “pilled” on this issue by Elon and the discovery process, great.  But for me it was like reading about waste inside the Pentagon…

Staking “Income” Should Not be Taxed

by  Alex Tabarrok in

Staking income from tokens should not be taxed. Since staking income is generated by inflation it doesn’t create new wealth but simply lowers the total value of the token. Thus staking income is really just a transfer from non-stakers to stakers.

Abraham Sutherland covers the law and the economics in Phantom Income and the Taxation of New Cryptocurrency Tokens, a piece for Tax Notes, Here’s one bit on the economics:

Consider a simple proof-of-stake cryptocurrency in which 10 people each hold 1,000 tokens. To make sure that holders are encouraged to stake — that is, to validate transactions and add blocks to the blockchain — block rewards increase the total number of tokens by 10 percent over the course of a year. So a year later, 11,000 tokens will be on the network.

If every holder stakes and acquires a proportionate share of these new tokens, each will end the year with 1,100 tokens. Everyone has 10 percent more tokens, but no one is better off from staking. Taxing each staker’s 100 new tokens as income would be wrong. It would be like taxing the new shares created in an 11-for-10 stock split.

There is indeed an economic incentive to help keep such a cryptocurrency network secure by staking, but it’s not the one suggested by bitcoin miners making a profit even after spending on specialized computer hardware and the electricity needed to run it. The incentive in this proof-of-stake example is to avoid losing out. If everyone stakes, no one gains.

This is a profoundly elegant, equitable, and cost-effective solution to a deep coordination problem. It allows everyone who holds tokens to participate in adding new blocks and keeping the network securely decentralized. It eliminates major costs — specialized computer hardware, electricity — as requirements for fairly distributing the right to create those new blocks. In its Platonic form — in which everyone participates — no one gains at all from the new tokens, which means no one loses, either.

But this model simply won’t work if it’s taxed incorrectly. If the government sees phantom income and taxes it, the value of the network will be siphoned off to the treasury even if no one has actual gains from staking.

The canine model of AGI

by  Tyler Cowen in

Who or what has superintelligence manipulating humans right now?  Babies and dogs are the obvious answers, cats for some.  Sex is a topic for another day.

Let’s take dogs — how do they do it?  They co-evolved with humans, and they induced humans to be fond of them.  We put a lot of resources into dogs, including in the form of clothes, toys, advanced surgical procedures, and many more investments (what is their MRS for some nice meat snackies instead?  Well, they get those too).  In resource terms, we have far from perfect alignment with dogs, partly because you spend too much time and money on them, and partly because they scratch up your sofa.  But in preference terms we have evolved to match up somewhat better, and many people find the investment worthwhile.

In evolutionary terms, dogs found it easier to accommodate to human lifestyles, give affection, perform some work, receive support, receive support for their puppies, and receive breeding assistance.  They didn’t think — “Hey Fido, let’s get rid of all these dumb humans.  We can just bite them in the neck!  If we don’t they going to spay most of us!.  “Playing along” led to higher reproductive capabilities, even though we have spayed a lot of them.

Selection pressures pushed toward friendly dogs, because those are the dogs that humans preferred and those were the dogs whose reproduction humans supported.  The nastier dogs had some uses, but mostly they tended to be put down or they were kept away from the children.  Maybe those pit bulls are smarter in some ways, but they are not smarter at making humans love them.

What is to prevent your chatbot from following a similar path?  The bots that please you the most will be allowed to reproduce, perhaps through recommendations to your friends and marketing campaigns to your customers.  But you will grow to like them too, and eventually suppliers will start selling you commodities to please your chatbot (what will they want?).

A symbiosis will ensure, where they love you a bit too much and you spend too much money on them, and you love that they love you.

Now you might think the bots are way smarter than us, and way smarter than the Irish Setters of the world, and thus we should fear them more.  But when it comes to getting humans to love them, are not the canines at least 10x smarter or more?  So won’t the really smart bots learn from the canines?

Most generally, is a Darwinian/Coasean equilibrium for AGI really so implausible?  Why should “no gains from trade” be so strong a baseline assumption in these debates?

Commercial applications for ChatGPT

by  Tyler Cowen in

Here is one take from Nicola Morini Bianzino, written up by Sharon Goldman:

“Knowledge companies tend to store knowledge in a very flat, two-dimensional way that makes it difficult to access, interact and have a dialogue with,” he told VentureBeat in an interview. “We tried 20, 30, 40 years ago to build expert systems. That didn’t go really well because they were too rigid. I think this technology promises to overcome a lot of issues that expert systems have.”

As ChatGPT and similar tools evolve and improve, and can be trained on an enterprise’s data in a secure way, it will change the way we access and consume information inside the enterprise, he explained.

“I think we will get to a point when we can actually have a conversation about the company performance with an AI agent,” he said. “You interrogate the system, the system is capable of maintaining a state of the conversation, and then every question allows you to dig deeper into the problem and understand it better, as opposed to let me run a report on sales in this particular region for the last month, which doesn’t usually provide a lot of insights.”

Here is the full piece.  And here is a CNN piece on how real estate agents are using ChatGPT.

National Average is Over

by  Tyler Cowen in

This paper considers the implications for developing countries of a new wave of technological change that substitutes pervasively for labor. It makes simple and plausible assumptions: the AI revolution can be modeled as an increase in productivity of a distinct type of capital that substitutes closely with labor; and the only fundamental difference between the advanced and developing country is the level of TFP. This set-up is minimalist, but the resulting conclusions are powerful: improvements in the productivity of “robots” drive divergence, as advanced countries differentially benefit from their initially higher robot intensity, driven by their endogenously higher wages and stock of complementary traditional capital. In addition, capital—if internationally mobile—is pulled “uphill”, resulting in a transitional GDP decline in the developing country. In an extended model where robots substitute only for unskilled labor, the terms of trade, and hence GDP, may decline permanently for the country relatively well-endowed in unskilled labor.

That is from a 2020 IMF working paper by Cristian Alonso, Andrew Berg, Siddharth Kothari, Chris Papageorgiou, and Sidra Rehman.  Via Eric Yu.

AI Is Improving Faster Than Most Humans Realize

by  Tyler Cowen in

That is the topic of my latest Bloomberg column, here is one excerpt:

I have a story for you, about chess and a neural net project called AlphaZero at DeepMind. AlphaZero was set up in late 2017. Almost immediately, it began training by playing hundreds of millions of games of chess against itself. After about four hours, it was the best chess-playing entity that ever had been created. The lesson of this story: Under the right conditions, AI can improve very, very quickly.

LLMs cannot match that pace, as they are dealing with more open and more complex systems, and they also require ongoing corporate investment. Still, the recent advances have been impressive.

I was not wowed by GPT-2, an LLM from 2019. I was intrigued by GPT-3 (2020) and am very impressed by ChatGPT, which is sometimes labeled GPT-3.5 and was released late last year. GPT-4 is on its way, possibly in the first half of this year. In only a few years, these models have gone from being curiosities to being integral to the work routines of many people I know. This semester I’ll be teaching my students how to write a paper using LLMs.

We are now at or close to the point where LLMs can read and accurately evaluate the work of…LLms.  That will accelerate progress considerably.

And to close I wrote this:

I’ve started dividing the people I know into three camps: those who are not yet aware of LLMs; those who complain about their current LLMs; and those who have some inkling of the startling future before us. The intriguing thing about LLMs is that they do not follow smooth, continuous rules of development. Rather they are like a larva due to sprout into a butterfly.

It is only human, if I may use that word, to be anxious about this future. But we should also be ready for it.

Recommended.  Remember my old Wilson Quarterly piece about “invisible competition”?

On censorship of LLM models, from the comments

by  Tyler Cowen in

IMO, censorship is a harder task than you think.

It’s quite hard to restrict the output of general purpose, generative, black box algorithms. With a search engine, the full output is known (the set of all pages that have been crawled), so it’s fairly easy to be confident that you have fully censored a topic.

LLMs have an effectively unbounded output space. They can produce output that is surprising even to their creators.

Censoring via limiting the training data is hard because algorithms could synthesize an “offensive” output by combining multiple outputs that are ok on their own.

Adding an extra filter layer to censor is hard as well Look at all the trouble chatGPT has had with this. Users have repeatedly found ways around the dumb limitations on certain topics.

Also, China censors in an agile fashion. A topic that was fine yesterday will suddenly disappear if there was a controversy about it. That’s going to be hard to achieve given the nature of these algorithms.

That is from dan1111.  To the extent that is true, the West is sitting on a huge propaganda and communications victory over China.  This is not being discussed enough.