ChiMan

No revolution revolutionizes in the way planned or hyped. Instead, once triggered, it emerges mostly unplanned. It unfolds as the product of human action but not of human design.

Individuals will use LLMs as they see fit in their work. They'll make a lot of mistakes because they don't see their own work clearly and because they don't really understand the outcomes others care about. Executives who make their money with their mouths will, of course, endlessly open their mouths, even as they don't understand the tech because they never really use any tech beyond email, text, and phones. And so on.

Nobody knows how LLMs will ultimately shake out. But even five minutes of using Claude makes it hard to imagine all white-collar work will continue as in the past.

I stopped the mid-morning bacon a few years ago and have put on 15 pounds since. A lot of variables, blah, blah, blah. I know. But still. No bacon, fewer eggs, and 15 pounds heavier, despite 10 - 20 more miles per week of running. Just sayin'.

Except that three saturated-fat-laden strips of bacon for breakfast prevents the need for lunch, leaving my daily caloric intake lower than if I had eaten a bagel for breakfast and been hungry at lunch. The longer period between meals--snack-free--afforded by the bacon also means that post-meal insulin (the energy-storing hormone) has a chance to drop off, leaving glucagon (the energy-burning hormone) a path for liberating stored energy to power physical and mental activity.

It seems the eggs improve the sleep, because after poor sleep, eggs the following day seem to have an effect on breaking the poor-sleep doom loop (which is: poor sleep elevates cortisol, adrenaline, heart rate, junk-food cravings, and the need for caffeine--all of which contribute to another night of bad sleep). Not sure if it's the choline, which eggs have a lot of, or the egg-generated satiety that helps prevent eating too much. I think it may be the choline because when I'm otherwise eating high-choline foods such as red meat, my sleep is also better.

Quality of this study aside, and n of 1 here, my own state of mind, clarity of thought, and sleep are all noticeably better when I'm eating 2 or 3 eggs a day, 3 to 5 days a week. (I might go 7 days, but an independent value placed on dietary variety prevents that--perhaps foolishly when I notice what I'm eating on off days instead of the eggs.)

Regardless, this whole eggs-are-evil thing has probably done more to harm the health of Westerners than any other dietary advice, with the possible exception of the fat-is-evil nonsense.

The monks likely have the time to think about implementation, and feeling like they’re part of an institution that transcends them and that they value for its own sake, they likely have an incentive to invest effort into maintaining and improving it.

Both of these are unlike, say, corporate environments, where the core work uses up almost all available time and where most people are looking mostly to extract something from the organization.

The rule of law isn’t based on the state’s monopoly on violence. Its entire purpose is to protect citizens from state violence by subordinating it to law adjudicated in courts.

Don't bore your co-workers (or others) with descriptions of your dreams, and don't throw a computer's dreams (AI chat logs) at them either.

Beautiful. An ode and axiom for our new age.

Covid-era over-hiring was never going to end well. If there’s any silver lining to this, it’s that with the AI tools used to AI-wash the dismissals, it’s easier than ever to bone up on economics and why it’s a bad idea to hire people without a clear view of the economic margins along which the employee will be profitable for the firm.

I wouldn’t count on reduced drudgery. The assembly line automated many movements needed for manufacturing. But which work involved more drudgery—-craftsman-style car production or standing on an assembly line at Ford?

With any new technology, subsequent drudgery depends on the technology, its concomitant economics, and the imagination of the people using it.

It also seems possible that clever, young outsiders will, by creatively employing AI, displace senior devs. Because unlike with law and accounting, novices in software development can generate a lot of their own reps at low cost (mostly just their time, which will be cheap compared to senior devs).

Another possibility could be using junior devs (with AI) to come up with enormous numbers of fresh ideas and minimally working programs, with the promising ones then worked on with senior devs.

Most fundamentally, because AI lowers the cost of testing out ideas and potential software products, it seems like a tool to test notions of what you should build or continue to offer customers in the first place.

Common sense and experience inform my theory of good sleep: Pitch black, stone quiet, with noise limited to pre-sleep audial approximations of the dream-like mental noise that precipitates sleep.

Yep. In general, when you see social solutions emerge in the wild from the free choices of participants, there is an underlying logic that’s doing important work. It’s easy to think you can improve over these natural solutions, but usually you can’t.

I recall working at this place that prided itself on starting meetings a minute early, as if trying to prove a point. The early starts did prove that, yes, you could get people to scramble, but not much else.

A norm of scheduling the “start” on the hour or half hour while really starting the meeting five minutes later also works. That way, attendees have an opportunity to arrive “on time” and chat if they like, building relationships in the process. The freedom to arrive at any point during the first five minutes also helps to create the kind of ease that’s conducive to serious discussion. This second part is particularly important when power dynamics might otherwise derail real discussion.

Because five minutes of pure chitchat can feel excessive to some folks, though, a three-minute norm probably works better—-especially because the off-centeredness has the informal aesthetic that, again, forms a better backdrop for serious discussion.

You can’t afford to risk being weird anymore. Housing is too expensive. Too many weird steps in non-pecuniary directions and you’ll end up on the street or your mom’s couch.

You know the weakness of man from a mile away by the verbosity and volume of his "toughness."

Of course, everyone in the room has already read the same leadership tips, likely earning you plenty of eye rolls and detracting from the straightforward, honest cooperation and on-task communication that are the backbone of all successful teams and companies.

This gets to the heart of why we have copyrights. They’re not to make writers rich. They’re to make us all rich with the content they produce.

The modern abuse of copyrights by the likes of Disney does not negate this otherwise wonderful institution.

Half the value of a todo app lies in the act of writing tasks down in the first place, and the other half lies in reading, revisiting, and revising them. In this, a .txt file works as well as anything, and is much cheaper. For me, a proud skinflint, the low cost has of a .txt file a side effect: I’m tickled to use it, reinforcing the memory that any todo app is meant to create.

Seems like a market has emerged for a LinkedIn-like service that does everything LinkedIn does, minus the public posts. Resumes and direct messages only.

The point is not to delegate. It’s to augment.

AI performs best in non-deterministic environments where highly extensive if slightly imperfect (or even hallucinatory) knowledge works just fine. When mapped onto today’s jobs, the fit feels less natural for high-level engineering than for “looser” tasks that would do well to be armed with wider knowledge. In other words, it seems like AI—or AI-armed humans—are more squarely aimed at executives.

Seems possible that one possible unintended consequence of AI could be a rebirth of the Web as something closer to what we knew. Because why use search at all for general inquiry when AI can satisfy much of that?

More critically, it’s not hard to imagine that, with AI-boosted boosted coding, a thousand bespoke search engines and other platforms being just around the corner, radically changing the economics of platform lock-in. When you can build your own version of Google Search with the help of AI and do the same with social media or any other centralizing Internet force, then platforms cease to be platforms at all. With AI, the challenges of self-hosting could become quite manageable as well. And while we’re at it, some version of the same, individual-centered computing economics on your own devices seems possible.

In these senses, it’s quite possible that Jobs’s vision of computing as extensions of individuals rather than individuals being extensions of computing is again at hand, with the magic of self-curated order from a chaotic Net not far behind.

The speed alone is sufficient for a local-first approach. The latency of any cloud software I’ve ever used is like constant sand in the gears of thinking. Although taking supplements that slow my thinking—essentially natural downers—do improve my experience with such software, the improved experience comes at the expense of IQ. Basically, you need to be a little slow and dumb for the software to work as intended.

This is nuts. Computers are supposed to enhance and enable thinking, not make you stupid. In this sense, cloud software is possibly the biggest fraud ever perpetrated on the paying, computer-using public.

For the love of God, please bring back my late 1990s and early 2000s brain-boosting computer experience.

The problem with that solution is that it would work too well, making it unattractive to lawmakers who need the issue to maintain their careers.

Major free-rider problem, this is—one, however, that property rights could protect against: You own your data. If the OpenAIs of the world want to ingest your data, they pay you for it.

Uses in AC and clothing seem like obvious use cases.

The more fundamental question is: If mass layoffs are so necessary, then who’s responsible for the unnecessary mass hiring? In any mass layoff, fire that person first. Get to the root of the problem.

With AI, I'd say white collar work should become less like factory work. Generative AI is non-deterministic, the opposite of factory work.

In four years, they’ll revisit the complexity. Facts may become simpler then.