Roblog

In a lengthy article that is my new gold standard for analysing the actual consequences of introducing AI into an industry, Justin Curl, Sayash Kapoor, and Arvind Narayanan explore the effects of AI on legal work.

Sayash and Arvind had previously authored AI as Normal Technology (also well worth reading), which argues that it’s a mistake by both boosters and doomsters to treat AI as some special category of thing that will lead to humanlike or superhuman intelligence; instead, we should treat it, and should remain in control of it, like any other technology. In particular, we should measure its impact by how it diffuses through society and industry, not by what it’s capable of in isolation, and we should expect that diffusion to take decades, not months.

This article is their attempt to apply the “AI as normal technology” thinking to a particular industry.

As a document- and language-focused industry, law has seemed ripe for LLM-based disruption; as an industry that often commands extremely high fees for its work, many people are desperate for that disruption to happen, and for legal work to become cheaper. That motivated reasoning is perhaps one of the reasons why law regularly tops the panic-inducing lists of “jobs that won’t exist in the future because of AI”.

The authors’ conclusions are much more nuanced. They look at regulatory concerns, which might inhibit AI’s progress into the industry; they look the fundamental dynamics of the industry, in this case the adversarial nature of common-law countries; they look at the impact of past technologies, which often failed to deliver the commodification of law that many people imagined.

AI, they conclude, won’t fix any of the structural problems faced by the legal industry and those who interact with it as clients, and may even magnify them; but it may be that the mere threat of AI disruption can be an external impetus to reform, reform that otherwise might have faltered through lack of will and coordination.

I think both the specific conclusions about law and the general framework outlined by the authors are enormously helpful contributions to the discourse around AI, and I’ll be trying to think along similar lines in my own work. #

The New York Times has developed a tool to download, transcribe, and summarise various right-wing podcasts, part of what they call the “manosphere”, in order to spot signs of division and discontent within Donald Trump’s base:

“When one of the shows publishes a new episode, the tool automatically downloads it, transcribes it, and summarizes the transcript. Every 24 hours the tool collates those summaries and generates a meta-summary with shared talking points and other notable daily trends. The final report is automatically emailed to journalists each morning at 8 a.m. ET. Currently, the tool is used by nearly 40 reporters across the newsroom.”

It’s a fascinating use of LLMs in the newsroom. Whether this ends up making bias worse, as the apparent consensus of grifters and racists becomes a signal that influences the Times’ own reporting, is a complicated but vital question. But this specific use-case is just one of many. The tool grew out of another, called Cheatsheet, which sounds essentially like a no-code tool that allows journalists to execute LLM-based workflows against custom data, and which has already led to significant investigative breakthroughs:

“The Initiatives Team started trialing other applications of LLMs to process large, messy datasets and file dumps on a case-by-case basis. Today, many of those live in a single spreadsheet-based tool. Reporters can drop datasets into Cheatsheet and then run different preset scripts and prompts. Each capability in the menu is known as a ‘recipe.’ Some of those recipes, like transcribing thousands of hours of video footage and summarizing transcriptions, are foundational to the Manosphere Report.

“Still in its beta, Cheatsheet has already been tested on about 300 users in the newsroom, with 50 of those being ‘really active users’, according to Seward. Right now, at least one new project is created in Cheatsheet every day. The tool has been used to investigate an election-interference group, to transcribe and translate Syrian prison records, and to find recent instances of Trump talking about Jan. 6. At times, Cheatsheet has even been used to take on more thorough historical analysis of podcasts.”

#

One of my favourite YouTube channels at the moment is @saintcavish, Christopher St. Cavish’s explorations of Chinese cooking. Christopher is no mere visitor, dipping in briefly and superficially to a vast and almost-unknowable cuisine; he has lived in Shanghai for over twenty years, and so is able to delve deep into culinary traditions and gain access to remarkable places, like this visit to a Michelin-starred fine-dining restaurant in Shenzhen.

The production values are perfect. There are no showy Netflix-style visuals or overly dramatic score. It’s just exceptionally clean, well-graded footage – shot by a single camera operator, Graeme Kennedy – of remarkable chefs cooking remarkable food, all with St. Cavish’s thoughtful commentary. Not one to watch while hungry. #

Creatives’ social media accounts are awash with lo-fi, analogue aesthetics, most of which are created digitally, often with cookie-cutter kits that undermine the whole idea of what “analogue” is supposed to represent. Elizabeth Goodspeed understands why:

“The practical reality is that most people no longer have the time, tools, or support to make fully analogue work, even if they want to. The creative infrastructure that would make it viable – materials access, slower timelines, financial stability – isn’t widely available. Designers and illustrators are stuck in a bind: analogue signals value, but digital is what’s feasible. The result is a kind of strategic mimicry. The market is looking for particular cues, and designers have to find a way to hit them. It doesn’t help that glossy, computer-made work can now be mistaken for AI either; clean, high-fidelity digital craft has become suspect by default, making handmade a safer choice. You can think of adding in fake ink splatters a bit like penciling in a beauty mark: an intentional imperfection done to signal authenticity, rather than the byproduct of a real nuisance.”

Rather than using these analogue cues for merely surface-level styling reasons, or to signal “this isn’t AI”, Elizabeth hopes that we can use this analogue fixation as part of a broader reckoning about how and why creative work is made, and to whose benefit:

“When analogue collapses into surface style, it stops applying pressure to how work is made and valued. But it doesn’t have to be that way. The Arts & Crafts movement, for instance, emerged alongside mass mechanisation and responded not with nostalgia, but with structural reorganisation. Designers and makers pushed back against the factory’s division of labour by reasserting continuity between thinking and making. Objects were produced slowly, often collaboratively, with an emphasis on material knowledge and visible decision-making. They were sold through guilds and exhibitions that foregrounded craft as labour, not just aesthetic, and delivered tangible financial benefits to the people who made them. There’s a version of today’s analogue fixation that could move in this direction.”

#

Via Matt Webb, from his broader (and fascinating) post about the nature of work and what the roboticisation of domestic labour might take from us, comes this topical and counter-intuitive nugget about Olympians. How is it that they bring themselves to do the presumably mad levels of hard work and training that are required to get to the tops of their games? Well, because they enjoy it:

“The very features of the sport that the ‘C’ swimmer finds unpleasant, the top level swimmer enjoys. What others see as boring – swimming back and forth over a black line for two hours, say – they find peaceful, even meditative, often challenging, or therapeutic. … It is incorrect to believe that top athletes suffer great sacrifices to achieve their goals. Often, they don’t see what they do as sacrificial at all. They like it.”

It’s from a 1989 ethnographic study on Olympic swimmers by Daniel Chambliss. The central premise is that levels of sports are discontinuous, and that, while putting in more and more effort can make you better within a particular level of performance, it isn’t what progresses you to the highest heights:

“Having seen that ‘more is better’ within local situations, we tend to extrapolate: if I work this hard to get to my level, how hard must Olympic swimmers work? If I sacrifice this much to qualify for the State Championships, how much must they sacrifice? We believe, extrapolating from what we learn about success at our own level, that they must work unbelievably hard, must feel incredible pressure, must sacrifice more and more to become successful. Assuming implicitly that stratification in sports is continuous rather than discrete (that the differences are quantitative) we believe that top athletes do unbelievable things. In short, we believe that they must be superhuman.”

Neither talent nor hard work are good explanations for their success. Instead, Chambliss argues, the cause is more qualitative than quantitative; the extremely successful do different things rather than more things, and keep doing them to a habitual and mundane degree.

“After three years of field work with world-class swimmers… I wrote a draft of some book chapters, full of stories about swimmers, and I showed it to a friend. ‘You need to jazz it up,’ he said. ‘You need to make these people more interesting. The analysis is nice, but except for the fact that these are good swimmers, there isn’t much else exciting to say about them as individuals.’ He was right, of course. What these athletes do was rather interesting, but the people themselves were only fast swimmers, who did the particular things one does to swim fast. It is all very mundane. When my friend said that they weren’t exciting, my best answer could only be, simply put: that’s the point.”

#

Dan Davies has written about how many systems (including the British constitution) use what he terms a “good chap” regulatory system: i.e. they rely on the system being mostly populated by decent people who obey the rules, even though they don’t technically have to.

That has obvious flaws, like when someone (Boris Johnson, say, or more disastrously Donald Trump) comes along who decides not to be a good chap. But, Dan argues, there’s a certain advantage to the naïveté:

“It is always possible to break norms, if you’re really determined to. Everything does, in fact, depend on having people in positions of power who respect the rules of the game. The British ‘good chap’ system is just much more blatantly in your face about it.

“Which might account for the longevity of the Westminster system. It is incredibly fragile, but it’s obviously fragile, and in this way achieves a sort of paradoxical antifragility. In a ‘good chap’ system, when a bad chap shows up, all the good chaps know that they have to band together and oppose or get rid of them, because they know that there are no systemic constraints on badchappery. In a system that’s meant to be full of checks and balances, it is much easier for a kind of bystander effect to develop, where everyone waits for the system to protect itself without understanding that the system is just them.”

The contrast with the US in its current mode is striking; I definitely see that bystander effect, where people continually expect some kind of check or balance to take effect. It certainly hasn’t yet. #

A brilliant takedown of the menu icons in the latest version of MacOS, a version I’m still steadfastly refusing to upgrade to, from Niki Prokopov.

Apple decided to do a thing that was both impossible and undesirable, then did it badly:

“In my opinion, Apple took on an impossible task: to add an icon to every menu item. There are just not enough good metaphors to do something like that.

“But even if there were, the premise itself is questionable: if everything has an icon, it doesn’t mean users will find what they are looking for faster.

“And even if the premise was solid, I still wish I could say: they did the best they could, given the goal. But that’s not true either: they did a poor job consistently applying the metaphors and designing the icons themselves.”

In referencing the Human Interface Guidelines from 1992, Niki makes a point that many people would do well to remember:

“…is an interface manual from 1992 still relevant today? Haven’t computers changed so much that entirely new principles, designs, and idioms apply?

“Yes and no. Of course, advice on how to adapt your icons to black-and-white displays is obsolete. But the principles – as long as they are good principles – still apply, because they are based on how humans work, not how computers work.

“Humans don’t get a new release every year. Our memory doesn’t double. Our eyesight doesn’t become sharper. Attention works the same way it always has. Visual recognition, motor skills – all of this is exactly as it was in 1992.”

#