Stephen Wolfram visits RWRI 18 (Summer Workshop) [PRIVATE]

Workshop organized by the Real World Risk Institute. The workshop is an intense 10-day online program, and the 18th edition took place from July 10-21, 2023.

This video discusses the capabilities and limitations of large language models like GPT, the challenges of setting constraints on AI systems, and the potential risks and consequences of AI decision-making. The video talks about:

  1. The concept of a “stochastic parrot” in language processing and machine learning.
  2. How language processing systems like GPT use data from the web to generate responses.
  3. Attempts to “trick” GPT with questions requiring nuanced understanding.
  4. The simple operation of GPT in predicting the next word in a sequence.
  5. The use of language models as a new interface to computers.
  6. The integration of GPT with Wolfram Alpha for computations and informed responses.
  7. The similarity between writing good prompts for GPT and expository writing.
  8. The training data for GPT, includes nonsense, fiction, and factual information.
  9. The problem of the “self-licking lollipop” in information sources.
  10. The concept of “necessarily human work” requires human choice and input.
  11. The potential for AI to make decisions and the challenges of setting constraints.
  12. A thought experiment called “promptocracy” for AI decision-making.
  13. The actuation layer of AI and the difficulty of setting constraints.
  14. The phenomenon of computational irreducibility and trade-offs in AI computation.
  15. The potential risks of AI decision-making and the need for understanding large language models.

The Most Intolerant Wins: The Dictatorship of the Small Minority

On Medium, Nassim explains how once an intransigent minority reaches a tiny percentage of the total population, the majority of the population will naturally succumb to their preferences:

The best example I know that gives insights into the functioning of a complex system is with the following situation. It suffices for an intransigent minority –a certain type of intransigent minorities –to reach a minutely small level, say three or four percent of the total population, for the entire population to have to submit to their preferences. Further, an optical illusion comes with the dominance of the minority: a naive observer would be under the impression that the choices and preferences are those of the majority. If it seems absurd, it is because our scientific intuitions aren’t calibrated for that (fughedabout scientific and academic intuitions and snap judgments; they don’t work and your standard intellectualization fails with complex systems, though not your grandmothers’ wisdom).

The main idea behind complex systems is that the ensemble behaves in ways not predicted by the components. The interactions matter more than the nature of the units. Studying individual ants will never (one can safely say never for most such situations), never give us an idea on how the ant colony operates. For that, one needs to understand an ant colony as an ant colony, no less, no more, not a collection of ants. This is called an “emergent” property of the whole, by which parts and whole differ because what matters is the interactions between such parts. And interactions can obey very simple rules. The rule we discuss in this chapter is the minority rule.

The minority rule will show us how it all it takes is a small number of intolerant virtuous people with skin in the game, in the form of courage, for society to function properly.

This example of complexity hit me, ironically, as I was attending the New England Complex Systems institute summer barbecue. As the hosts were setting up the table and unpacking the drinks, a friend who was observant and only ate Kosher dropped by to say hello. I offered him a glass of that type of yellow sugared water with citric acid people sometimes call lemonade, almost certain that he would reject it owing to his dietary laws. He didn’t. He drank the liquid called lemonade, and another Kosher person commented: “liquids around here are Kosher”. We looked at the carton container. There was a fine print: a tiny symbol, a U inside a circle, indicating that it was Kosher. The symbol will be detected by those who need to know and look for the minuscule print. As to others, like myself, I had been speaking prose all these years without knowing, drinking Kosher liquids without knowing they were Kosher liquids.

Figure 1 The lemonade container with the circled U indicating it is (literally) Kosher.

Criminals With Peanut Allergies

A strange idea hit me. The Kosher population represents less than three tenth of a percent of the residents of the United States. Yet, it appears that almost all drinks are Kosher. Why? Simply because going full Kosher allows the producer, grocer, restaurant, to not have to distinguish between Kosher and nonkosher for liquids, with special markers, separate aisles, separate inventories, different stocking sub-facilities. And the simple rule that changes the total is as follows:

A Kosher (or halal) eater will never eat nonkosher (or nonhalal) food , but a nonkosher eater isn’t banned from eating kosher.

Or, rephrased in another domain:

A disabled person will not use the regular bathroom but a nondisabled person will use the bathroom for disabled people.

Granted, sometimes, in practice, we hesitate to use the bathroom with the disabled sign on it owing to some confusion –mistaking the rule for the one for parking cars, under the belief that the bathroom is reserved for exclusive use by the handicapped.

Someone with a peanut allergy will not eat products that touch peanuts but a person without such allergy can eat items without peanut traces in them.

Which explains why it is so hard to find peanuts on airplanes and why schools are peanut-free (which, in a way, increases the number of persons with peanut allergies as reduced exposure is one of the causes behind such allergies).

Let us apply the rule to domains where it can get entertaining:

An honest person will never commit criminal acts but a criminal will readily engage in legal acts.

Let us call such minority an intransigent group, and the majority a flexible one. And the rule is an asymmetry in choices.

I once pulled a prank on a friend. Years ago when Big Tobacco were hiding and repressing the evidence of harm from secondary smoking, New York had smoking and nonsmoking sections in restaurants (even airplanes had, absurdly, a smoking section). I once went to lunch with a friend visiting from Europe: the restaurant only had availability in the smoking sections. I convinced the friend that we needed to buy cigarettes as we had to smoke in the smoking section. He complied.

Read the rest of the article.

Nassim Shares Chart Comparing Outcomes with Clinton as President Versus Trump

Recently on twitter, Nassim  shared this chart comparing outcomes with Clinton as president versus Trump.  In response to a question about Sanders versus Trump, he tweeted: “Bernie v. Trump I go Bernie because his domestic policies won’t go through and his foreign policies are clearer.”

comparing-outcomes-with-hillary-clinton-vs-donald-trumpHe also followed up with these tweets: “Hillary and Bush have done the most to harm minority populations in the Levant & Iraq since Genghis Khan.” “2/What I mean by rigorous is not making assumptions/ certainties when there is uncertainty. Shillary offers certainties, Trump fewer ones.”

EconTalk with Russ Roberts: Taleb on Skin in the Game

econ-talk-library-of-economics-and-liberty-nassim-talebNassim Taleb of NYU-Poly talks with EconTalk host Russ Roberts about his recent paper (with Constantine Sandis) on the morality and effectiveness of “skin in the game.” When decision makers have skin in the game–when they share in the costs and benefits of their decisions that might affect others–they are more likely to make prudent decisions than in cases where decision-makers can impose costs on others. Taleb sees skin in the game as not just a useful policy concept but a moral imperative. The conversation closes with some observations on the power of expected value for evaluating predictions along with Taleb’s thoughts on economists who rarely have skin in the game when they make forecasts or take policy positions.

Website: http:// www. econtalk. org/ archives/ 2013/ 09/ taleb_ on_ skin_i. html
Direct Link (mp3): http:// files. liberty fund. org/ econtalk / y2013 /Taleb skin. mp3
Link to the Paper: http:// papers. ssrn. com/ sol3/ papers.cfm? abstract_id= 2298292