AI passes the restaurant review Turing test

by  Tyler Cowen in

Surprisingly easy, it turned out. In a series of experiments for a new study, Kovács found that a panel of human testers was unable to distinguish between reviews written by humans and those written by GPT-4, the LLM powering the latest iteration of ChatGPT. In fact, they were more confident about the authenticity of AI-written reviews than they were about human-written reviews.

Here is the full story, via Sarah Jenislawski.

Money for Blood and Short Term Jobs

by  Alex Tabarrok in

The excellent Tim Taylor on a new paper on plasma donation:

John M Dooley and Emily A Gallagher take a different approach in “Blood Money: Selling Plasma to Avoid High-Interest Loans” (Review of Economic Studies, forthcoming, published online May2, 2024; SSRN working paper version here). They are investigating how the opening of a blood plasma center in an area affects the finances of low-income individuals. As background, they write:

“Plasma, a component of blood, is a key ingredient in medications that treat millions of people for immune disorders and other illnesses. At over $26 billion in annual value in 2021, plasma represents the largest market for human materials. The U.S. provides 70% of the global plasma supply, putting blood products consistently in the country’s top ten export categories. The U.S. produces this level of plasma because, unlike most other countries, the U.S. allows pharmaceutical corporations to compensate donors – typically about $50 per donation for new donors, with rates reaching $200 per donation during severe shortages. The U.S. also permits comparatively high donation frequencies: up to twice per week (or 104 times per year)…”

Not too surprisingly, plasma donors tend to be young and poor and they use plasma donation to substitute away from non-bank credit like payday loans.

I am struck by Tim’s thoughts on how this connects with labor markets and regulation:

…I find myself thinking about the financial stresses that many Americans face. Being paid a few hundred dollars for a series of plasma donations isn’t an ideal answer. Neither is taking out a high-interest short-term loan; indeed, taking out a loan at all may be a poor idea if you aren’t expecting to have the income to pay it back. In the modern US economy, hiring someone for even a short-term job involves human resources departments, paperwork, personal identification, bookkeeping, and tax records. These rules have their reasons, but the result is that finding a short-term job that pays for a few days work isn’t simple, even if though most urban areas have a semi-underground network of such jobs.

Roger Miller’s classic 1964 song, “King of the Road,” tells us that “two hours of pushin’ broom/ Buys an eight by twelve four-bit room.” Even after allowing for a certain romancing of the life of a hobo in that song, the notion that a low-income person can walk out the door and find an two-hour job that pays enough to solve immediate cash-flow problems–other than donating plasma–seems nearly impossible in the modern economy.

My Conversation with the excellent Michael Nielsen

by  Tyler Cowen in

Here is the audio, video, and transcript.  Here is the episode summary:

Michael Nielsen is scientist who helped pioneer quantum computing and the modern open science movement. He’s worked at Y Combinator, co-authored on scientific progress with Patrick Collison, and is a prolific writer, reader, commentator, and mentor. 

He joined Tyler to discuss why the universe is so beautiful to human eyes (but not ears), how to find good collaborators, the influence of Simone Weil, where Olaf Stapledon’s understand of the social word went wrong, potential applications of quantum computing, the (rising) status of linear algebra, what makes for physicists who age well, finding young mentors, why some scientific fields have pre-print platforms and others don’t, how so many crummy journals survive, the threat of cheap nukes, the many unknowns of Mars colonization, techniques for paying closer attention, what you learn when visiting the USS Midway, why he changed his mind about Emergent Ventures, why he didn’t join OpenAI in 2015, what he’ll learn next, and more. 

And here is one excerpt:

COWEN: Now, you’ve written that in the first half of your life, you typically were the youngest person in your circle and that in the second half of your life, which is probably now, you’re typically the eldest person in your circle. How would you model that as a claim about you?

NIELSEN: I hope I’m in the first 5 percent of my life, but it’s sadly unlikely.

COWEN: Let’s say you’re 50 now, and you live to 100, which is plausible —

NIELSEN: Which is plausible.

COWEN: — and you would now be in the second half of your life.

NIELSEN: Yes. I can give shallow reasons. I can’t give good reasons. The good reason in the first half was, so much of the work I was doing was kind of new fields of science, and those tend to be dominated essentially, for almost sunk-cost reasons — people who don’t have any sunk costs tend to be younger. They go into these fields. These early days of quantum computing, early days of open science — they were dominated by people in their 20s. Then they’d go off and become faculty members. They’d be the youngest person on the faculty.

Now, maybe it’s just because I found San Francisco, and it’s such an interesting cultural institution or achievement of civilization. We’ve got this amplifier for 25-year-olds that lets them make dreams in the world. That’s, for me, anyway, for a person with my personality, very attractive for many of the same reasons.

COWEN: Let’s say you had a theory of your collaborators, and other than, yes, they’re smart; they work hard; but trying to pin down in as few dimensions as possible, who’s likely to become a collaborator of yours after taking into account the obvious? What’s your theory of your own collaborators?

NIELSEN: They’re all extremely open to experience. They’re all extremely curious. They’re all extremely parasocial. They’re all extremely ambitious. They’re all extremely imaginative.

Self-recommending throughout.

Wednesday assorted links

by  Tyler Cowen

1. Rave review for Richard Flanagan’s Question.

2. Longer novels are more likely to win literary awards.

3. Typing Chinese on a QWERTY keyboard.

4. “ChatGPT is used most frequently in India, China, Kenya and Pakistan, with 75%, 73%, 69% and 62% of respondents, respectively, reporting using ChatGPT daily or weekly”  Link here.

5. Mark Koyama on Geoffrey Hodgson.

6. William Shoki on South Africa (NYT).  And Jeffrey Paller on the South African elections.

Monaco on the Marin Headlands

by  Alex Tabarrok in

The Dalmation Coast in Croatia, the Amalfi Coast in Italy and Monaco’s coast on the Mediterranean Sea are often found on lists of the most beautiful coastlines in the world. Here are some pictures. Hard not to agree. The fourth picture is of the Marin coastline near San Francisco. It’s also beautiful but is it obviously more beautiful than the other coastlines? Personally, I don’t think so. But one thing is different. Far fewer people are enjoying the Marin coast. Why? Because fewer people live there. Can something be beautiful if there is no one to see it?

There is something to be said for protecting natural wilderness but must we do so on some of the most valuable land in the world?

I agree with Market Urbanism, “Quite simply, we must build Monaco on the Marin Headlands.”

Hat tip to Bryan Caplan who makes the point about beauty in his excellent, Build, Baby, Build.

Croatia

 

Amalfi

Monaco

Marin:

What film or literature is useful for making sense of the AI moment?

by  Tyler Cowen

That is a reader query, I take it there is no point in my trotting through the obvious picks, starting with I, Robot and Her.  Sophisticates will ponder Stanislaw Lem’s Cyberiad, which very well understood the quirky and semi-religious potential in LLMs, even though he was writing in Communist Poland a very long time ago (those people loved to talk about cybernetics).

Have you ever pondered the 1994 Sandra Bullock movie Speed?  I think at this point it is not a spoiler to report “It revolves around a bus that is rigged by an extortionist to explode if its speed falls below 50 miles per hour.”  And yes, this is a Hollywood movie of the 1990s, so it does end in a kiss.

Here is a visual mapping of how science fiction has made sense of AI.  Am I neglecting any other non-obvious picks?

Urban design taken seriously?

by  Tyler Cowen

GPT-4o suggested:

“1. **Serendipity Corners**:

– Implement areas designed to encourage unexpected encounters. For example, interactive art installations or quirky features can serve as conversation starters. These could change frequently to keep the city dynamic, such as rotating sculptures or murals with interactive elements like touch-activated sound…

3. **Social Puzzles**:

– Install public games or puzzles that require collaboration from multiple people. This could be anything from giant chess boards to augmented reality treasure hunts that encourage teams to work together to solve clues scattered throughout the city.”

Claude’s answers were along broadly similar lines.  For an economic answer, how about “raise the city income tax on working”?  Love is not taxed, but work income is.  Furthermore, these days it is relatively difficult to strike up romantic relationships in many kinds of workplaces.   (Of course you would need offseting “stay in the city” subsidies, in balanced budget fashion.)  Taxing female education is another bad idea, but if the unconstrained goal is to increase the number of love matches that might work too.

What else?

Those circularity-inducing service sector jobs

by  Tyler Cowen in

Yes, in Tokyo women pay men in make-up to flatter them.  But what does the whole market look like?

Yamada Kurumi, a client, works at a brothel to earn enough money to visit the clubs, which she does about once a week. She had boyfriends in the past but finds hosts more exciting. She is unsure whether to seek an office job after graduating from college or to carry on with sex work, which pays better. “A lot of people start losing touch with friends once they get addicted to host clubs,” says Ms Yamada. “My host is already part of my everyday life…If I get a normal job, I probably won’t be able to see him any more. That scares me.”

Here is more from The Economist.

Deadly Precaution

by  Alex Tabarrok in

MSNBC asked me to put together my thoughts on the FDA and sunscreen. I think the piece came out very well. Here are some key grafs:

…In the European Union, sunscreens are regulated as cosmetics, which means greater flexibility in approving active ingredients. In the U.S., sunscreens are regulated as drugs, which means getting new ingredients approved is an expensive and time-consuming process. Because they’re treated as cosmetics, European-made sunscreens can draw on a wider variety of ingredients that protect better and are also less oily, less chalky and last longer. Does the FDA’s lengthier and more demanding approval process mean U.S. sunscreens are safer than their European counterparts? Not at all. In fact, American sunscreens may be less safe.

Sunscreens protect by blocking ultraviolet rays from penetrating the skin. Ultraviolet B (UVB) rays, with their shorter wavelength, primarily affect the outer skin layer and are the main cause of sunburn. In contrast, ultraviolet A (UVA) rays have a longer wavelength, penetrate more deeply into the skin and contribute to wrinkling, aging and the development of melanoma, the deadliest form of skin cancer. In many ways, UVA rays are more dangerous than UVB rays because they are more insidious. UVB rays hit when the sun is bright, and because they burn they come with a natural warning. UVA rays, though, can pass through clouds and cause skin cancer without generating obvious skin damage.

The problem is that American sunscreens work better against UVB rays than against the more dangerous UVA rays. That is, they’re better at preventing sunburn than skin cancer. In fact, many U.S. sunscreens would fail European standards for UVA protection. Precisely because European sunscreens can draw on more ingredients, they can protect better against UVA rays. Thus, instead of being safer, U.S. sunscreens may be riskier.

Most op-eds on the sunscreen issue stop there but I like to put sunscreen delay into a larger context:

Dangerous precaution should be a familiar story. During the Covid pandemic, Europe approved rapid-antigen tests much more quickly than the U.S. did. As a result, the U.S. floundered for months while infected people unknowingly spread disease. By one careful estimate, over 100,000 lives could have been saved had rapid tests been available in the U.S. sooner.

I also discuss cough medicine in the op-ed and, of course, I propose a solution:

If a medical drug or device has been approved by another developed country, a country that the World Health Organization recognizes as a stringent regulatory authority, then it ought to be fast-tracked for approval in the U.S…Americans traveling in Europe do not hesitate to use European sunscreens, rapid tests or cough medicine, because they know the European Medicines Agency is a careful regulator, at least on par with the FDA. But if Americans in Europe don’t hesitate to use European-approved pharmaceuticals, then why are these same pharmaceuticals banned for Americans in America?

Peer approval is working in other regulatory fields. A German driver’s license, for example, is recognized as legitimate — i.e., there’s no need to take another driving test — in most U.S. states and vice versa. And the FDA does recognize some peers. When it comes to food regulation, for example, the FDA recognizes the Canadian Food Inspection Agency as a peer. Peer approval means that food imports from and exports to Canada can be sped through regulatory paperwork, bringing benefits to both Canadians and Americans.

In short, the FDA’s overly cautious approach on sunscreens is a lesson in how precaution can be dangerous. By adopting a peer-approval system, we can prevent deadly delays and provide Americans with better sunscreens, effective rapid tests and superior cold medicines. This approach, supported by both sides of the political aisle, can modernize our regulations and ensure that Americans have timely access to the best health products. It’s time to move forward and turn caution into action for the sake of public health and for less risky time in the sun.

*Crooked Plow*

by  Tyler Cowen in

That is a recently translated Brazilian novel by Itamar Vieira Junior, set in Bahia.  It is better to read this one without any spoilers.  And I am pleased to announce that we have another great Latin American (and Brazilian) novel, worthy of entering the canon.

I haven’t seen a good ungated review of the book, as no one seems to care.  I did like this NYRoB (gated) review.  Further on the plus side, the book is also short and an easy read.