83. Information Work Lotteries 🔗

Sufficiently advanced information work is indistinguishable from a lottery ticket.

If you are enough degrees of abstraction and indirection from anything real, you have about the same chance of having say a million-dollar impact on reality as you do of winning a million dollar lottery ticket.

The people who seem to have been happiest through the Great Weirding are the ones who enjoy a relatively deterministic relationship between work and value. You put in effort here, and a predictable amount of value accrues in a predictable place over there.

This is a subtle source of angst. The indeterminacy of locus and extent of value from what you do, as measured by people actually saying they got value, weighted by your valuation of them as humans.

One of the reasons I like my idea of baseline "you have no obligation to be interesting or useful to the world" is that it forces you to ask whether whatever you do is worth doing entirely for its own sake, making any value accrual anywhere else pure upside that you don't need.

The moment you link your personal fulfilment to value accrual at an external locus, indeterminacy in the locus, the value itself, and the value to you of that locus, muddy the equation.

And to the extent you can't even observe it, it becomes impossible. Imagine if you were a writer with no feedback loop at all. Everything you wrote went into a blackhole with no indication at all, ever, of whether anyone even read it... would you still write?

Writing in fact throws you into deep end when you start. It will seem like screaming into the void with zero response. If you don't get enough out of it by yourself, you won't last long enough to see anyone else get anything out of it. So if you do it for others, you'll fail.

I'd say on a scale of 1 to 10, 10 being "pure lottery ticket," writing nonfiction on spec, with no pay, for general consumption in public is probably about an 8. Fiction is 9. Poetry is 9.5.

Interesting to think of apparent bullshit work (high abstraction/indirection) that's mainly powerpoints and other middle-management vagueware in this light. It's not so much bullshit as aggregated lottery work in a sort of mining pool.

In a 10k person org with say 8k individual contributors, 1980 managers, and 20 executive management, I think the bet is that in any given quarter, at least a few of the powerpoints made by the 1980 managers "pays off" and the benefit is socialized across the orgs.

Twitter is a good example of such a lottery system. Even when you have enough of a following that you can sort of predict how many likes/RTs a tweet will generally have, you can't always predict who will like/RT or when you'll get the random viral blow-up.

On certain kinds of tweets you can predict who will see and like/RT, but the algorithm adds enough of a lottery-like uncertainty that it's never really certain. This is why you have 3 basic tweeting styles: for yourself, for mutuals, for the numbers.

If all you care about is the number of likes and RTs, it means you don't really care about who is responding, which means you value every potential respondent equally low, so you need to aggregate a lot.

The other extreme is tweeting for a single senpai-notice-me muse.

I've only ever tweeted for myself. Once I get to know a mutual, I might @ them for specific things. Both the numbers stuff and senpai-notice-me styles leave me cold. Because they ARE both "cold" styles in their own way. Kinda clinically playing the lottery in a specific way.

Playing for numbers is like slots, but with some ability to pick out machines that return greater than 100% (the fortune cookie themes like personal development, startups, crypto...).

Playing senpai-notice-me is like blackjack. Skilled tactics like card counting are available.

Somebody asked me yesterday what I do with my time, and I said "waste it."

It's kinda true. To the extent that I spend most of my time in extremely lottery-like regimes of information work, I "waste" it.

I've made several years worth of my income via interactions that started on Twitter, so in some sense, all my tweeting is highly non-deterministic "work". But it's close enough to lottery that it doesn't feel like it. I can't with a straight face call it work.

Is it still a "target audience" if you don't actually target them? 🤔

In fact, the more you "target" the less of a lottery it is, which means you have to make it less abstract/indirect kinds of information work.

Many people don't get this. They write articles like "How AI can improve your business" and it reads like pablum and flops precisely because the abstraction level doesn't match the targeting precision. At that precision, you should write something like "PyTorch for Managers."

If you DO want to write something general and abstract about AI and business, it needs to be much less "targeted." Maybe something based on a metaphor, like "AI is the new oil" that will play much more unpredictably than an on-the-nose thing.

This lottery mindset is not the same as a "numbers game" like sales cold-calling or spray-and-pray resume broadcast. There you're pushing a concrete thing and uncertainty comes from competition or unknowns. Here the uncertainty is created by the abstraction of what you're doing.

You could call it transactional uncertainty versus causal uncertainty. In transactional uncertainty, you know how and why value can be created and just need to close the deal. In causal uncertainty, you don't know how an idea can actually create value.

Transactional uncertainty can be mitigated with better communication and improved filtration and targeting technology. Your precision is only limited by things outside of the work itself. Causal uncertainty is intrinsic to the work itself.

Other things being equal, the value of writing is fundamentally more uncertain the value of programming or cooking or vacuuming the carpet. And within wriiting, a philosophy essay has much more uncertainty in value than Ikea instructions.

The "information work" character and degrees of abstraction/indirection are often very subtle though. For eg. I'm trying to build a hobby-scale rover. In relation to any possible place this could have value for anyone else, it is much more abstract and indirect than my writing.

The actual rover, if I ever finish it, is of no use to anyone, not even me. It'll be a toy. The designs may be of use to someone who may go on to do interesting things. The thing is an extreme lottery ticket. Like an absurd 9.9/10.