Previously: Escape Velocity From Bullshit Jobs
They took our jobs! They took… several of our jobs?
“ChatGPT does like 80 percent of my job,” said one worker. Another is holding the line at four robot-performed jobs. “Five would be overkill,” he said.
The stories told in this Vice article (and elsewhere) are various people who found themselves able to do most job tasks much faster than they could previously do those same tasks. They use this to take on additional ‘full time’ jobs.
If the jobs involved are productive, this reflects large increases in total factor productivity that will continue to spread. That seems great.
Still, this multiple jobs reaction is a little weird. Seems worth exploring a bit more. Why are such people taking on multiple jobs, rather than doing better at one job?
A Question of Composition and Compensation
Simple math plus social dynamics. Multiple ‘full time’ jobs lead to much better pay.
In most salaried jobs, compensation is mostly dictated by social status and hierarchy. You are mostly paid what people with your title and position are paid.
A killer employee who can produce ten times as much work product of identical quality per hour, or superior work worth ten times as much – such as the standard ‘10x programmer’ – would be supremely lucky to earn double the pay of a worker of average skill.
This is a key reason why great employees are so valuable. Not only do you only manage and communicate with one person instead of ten, you save tons of money.
Why does it work this way? Compensation in jobs is inherently about comparisons, about social status, about hierarchy and about fairness norms. It is also about what can be justified to others, what is standard and what sounds ‘reasonable.’ If you tried paying your 10x employee five times what your 1x employees get paid, the 1x employees would revolt, the boss would think you were crazy and the funders would raise hell, you would resent that they’re paid more than you are, and so on. Also, they’d know the employee didn’t ‘need’ the money. Who do they think they are?
Our norms continuously push such dynamics towards equality.
Workarounds for Insufficient Pay Inequality
We do have several known existing solutions for this. I’ll consider the top three.
Option 1: The default reward is that by doing amazing work you would ‘get promoted’ into a different job, where you would have different tasks. Where you are likely not 10x the standard. One of those tasks will be management of other employees, which you likely hate and do not want to do. This then gives social justification for increased compensation, but destroys value and your experience. It will probably take many promotions to triple your compensation. Getting those promotions will likely require you doing battle inside a moral maze and be unlinked to your newfound production.
It is easy to see why someone getting a supercharge to their productivity from GPT-4 would be uninterested in climbing the corporate ladder.
Option 2: One can break free from salary or even hourly pay entirely, as some fields and professions allow. If you are for example a 10x salesman, artisan, trader, merchant, tournament competitor, poker player or author, you keep the extra production. There is no need to split your attentions to justify increasing your compensation.
Option 3: If you want to make the really big bucks, you do not want a job. What you want is equity. You want skin in the game. That is The Way.
I highly recommend to anyone who is capable and motivated that you want to take option three to the greatest extent possible. Early employee is good. Founder is better.
It is not simple, easy or safe.
Founding a company or running a business requires taking on a wide variety of other problems and tasks, and taking on a lot of risk. It is not for everyone, even among those who are highly productive and self-motivating.
The Overemployment Option
Often none of the three above options will be viable. Promotions don’t offer much, your profession has equality norms for compensation and you don’t see a good path to equity. Yet your productivity is now very high.
Thus:
Option 4: You take multiple jobs. It is much easier to do 10 hours of work for each of three jobs at $250k each than it is to do 40 hours of work and get them to pay you $750k.
This is even more true if the large positive productivity shock is available to others as well, and uses methods others might copy, frown upon or ban. If you suddenly 5x your production people are going to want to know why.
Note both the similarities and differences to those who work two lower-end jobs in order to make ends meet, where various norms prevent earning more pay from only one job, no matter how many hours they could put in or how productive they could be.
Should You Do This?
You should absolutely do the part where you get massive productivity gains, if your profession of choice allows massive productivity gains.
The question is what to do with those gains. When is multiple jobs the way?
If you need to take on multiple jobs to pay the bills, consider trying to find better paying work, and ask whether you’re spending more than you need to, and so on, but certainly lack of solvency is a good reason to go down the multiple jobs path at least for a time.
Otherwise, assuming you can pay the bills without additional jobs, my take is:
Prioritize getting skin in the game. Equity rules everything around you.
Only then, ask whether what you’re already doing effectively ‘caps out’ in time.
Get sleep, keep sane, have a social life, learn, explore and so on.
Even if your primary job can’t provide skin in the game, taking on a second project is a great opportunity to pursue skin in the game.
If you still have enough spare cycles and you can’t do better, sure, more jobs.
What Next?
What happens when more and more people get such large boosts in productivity?
This could go in several different directions, which can be combined.
Acceptance. Employers increasingly look the other way, actively not minding if you take multiple jobs or even outsource your work. What matters are the results, and how much it costs to get those results.
Recalibration. Employers see the increase in productivity, demand lots more production from each worker, so it becomes impossible for all but a handful of people to hold down multiple jobs or only put in a few hours.
Dejobization. If what matters is the work, perhaps stop thinking about people as having ‘jobs’ in such contexts. More work transitions to contracts, commissions, commerce and per-result payments.
Bullshitization. If the purpose of jobs is to ensure that people have exactly one job, no more and no less, or the purpose of a job is to know that there is a person dedicated to a particular thing, then the work must expand, through ever-increasing amounts of bullshit, until the job takes the correct number of hours. That could take the form of having to be in the office, or banning the use of LLMs. It could take the form of endless meetings, although beware whisper transcriptions and summary requests being used to ignore meetings. It can mean more paperwork, more forms, more things designed to be difficult to automate, and so on. The possibilities are endless.
Regulation. Crack down on the bastards. Keep a registry of who works where and flag anyone with multiple jobs, destroy the social permissions and force them to sit idle the rest of the week or at least do something that isn’t shaped like a job. Plausible this is good, since jobs are generally not the most productive use of time.
What happens to wages and jobs? Almost anything is possible here, almost no matter the scenario, both in a particular area or in general.
Paradise. Productivity rises, more good things, more surplus, higher real wages and returns to capital, more demand for everything good, lots of things now worth doing. Things are moving fast, you do better if you adapt, cost disease still ensures everyone gets paid well.
Dystopia. Productivity rises, less jobs are needed, demand doesn’t rise that much, too many workers chase too few jobs, wages crash. Everyone without capital is scrambling to stay above water, as even if you have a ‘safe’ job there are lots of people who enter that job after losing their old one.
Redistribution. Dystopian dynamics by default, but we institute a UBI, or alternatively some sort of jobs program (even if that program is ‘require a lot of bullshit’) and this results in sufficient redistribution that everyone is mostly fine.
Would be remiss to not also mention, of course:
Slow Motion Doom. Central example: AIs soon do almost all remote jobs better than humans, anyone using a human gets outcompeted, then anyone using a not-fully-unleashed AI gets outcompeted, then the unleashed AIs rapidly get control of the future, humans have steadily less real resources over time and soon die out, even if the AIs ‘respect property rights’ in a way that humans never actually did.
Quick Doom. In the end, none of these dynamics mattered. Foom.
There will be some areas where there isn’t that much latent demand waiting to be induced, and where employment will collapse. There will be other areas where more productivity means more production, I expect software engineering to be in that space for a while although people disagree on this. And there will be areas where new jobs come to exist, or our newfound surplus induces demand despite not much change in sector productivity.
Overall, I remain an optimist about job impacts, especially for those paying attention and willing to adapt. That includes realizing that a job may not be what you need.
There’s a model of white-collar employment I think is missing here. (I also thought it was missing at some points in the Moral Mazes sequence, but never got around to writing it down then).
The model is underutilized employees as option value.
Imagine yourself as a manager, running a small team at some company somewhere.
Most weeks, your team has 40 hours of work to do.
Every so often, there is a crisis. Perhaps your firm’s product is scheduled to release in 2 weeks when a regulator suddenly dumps 500 pages of compliance questions on you and refuses to approve your product until they are answered. Perhaps a security flaw is discovered in a major library you use and you need to refactor your whole codebase. Perhaps someone makes a mistake and your firm’s biggest client is ticked off. In any case, you have 200 hours of work to do that week, and if it is not done that will be Very Bad.
How many employees should you hire?
You could hire one employee. This employee would be busy but not unmanageably so most weeks...and then as soon as something went wrong, there would be a disaster.
So instead you hire perhaps four employees. Most weeks they each have 10 hours of work to do, and spend the rest of the time chatting around the water cooler or playing solitaire on work computers or whatever. And when a fire drill happens, you get them to drop the solitaire games and maybe put in a few extra hours and you have enough people to handle it.
The thing you are buying with these three extra employees is not the 10 hours they each work in a typical week. It is them being around and available and familiar with the job when something goes wrong.
-------END OF MODEL, BEGINNING OF ARGUMENT------
Many people seem to me to be operating on a model something like this:
Under that model, if you can do your job in 10 hours a week with GPT that is great! This high productivity should lead to some improvements, either in you having more free time, or in you being able to get more jobs and make more money.
Under my model, of course you can do your job in 10 hours most weeks. This has nothing to do with GPT! Your job can be done in 10 hours most weeks because most weeks nothing is on fire.
But if you have two employees:
Alice shows up at the office, does 10 hours of work, and plays solitaire for 30 hours.
Bob works from home, does 10 hours of work, and then switches to Job #2.
Bob is actually worth much less as an employee. Because when something goes wrong, you can grab Alice, call in your option, get her to stop playing solitaire, and have her do a bunch of work that was needed...but when you try to call your option on Bob, he’ll be on a call with his manager from Job #3. Your actual binding constraint is how much work needs to get done during a crisis, and Bob contributes very little to that.
A related bit is that you can’t generally respond to occasional weeks when you need 200hr of work by bringing in new people, or people from other areas of the company. You need people who already understand your systems, understand the general shape of the work (at least in normal times), and know how to work together.
You also need to be able to handle losing an employee. Even if you could get along fine with a single competent highly utilized person, if everything depends on them and they quit you’re in massive trouble. Much less so if you have three people who could each do all the work on their own.
There are often large gains to be found in finding lower priority work for people to be doing that replaces the solitaire, though pushing too hard here risks losing people.
This can also be solved by having multiple people work on multiple projects each. For example, you have 4 people and 4 projects, everyone works 25% on each project. The idea is that some projects have higher priority, some have lower, so when you need extra power on the important project, you can tell all 4 to prioritize it and give it 50% or 90%.
And it sucks a lot. Typically because each project has a separate project manager, and even if one is low-priority, its manager typically insists that you keep working on it at the usual speed, even when you keep getting extra tasks on the high-priority one.
I like this model, and it usefully describes a reality that I sometimes see in the workplace.
However, I claim that this is not the model that anyone’s employers are using. The decisions made around hiring budgets, as they have been explained to me, plan for the standard case: in your example, 40 hours of work per week and so one employee gets hired. In the event of crisis the employee is expected to put in extra time (which costs nothing in the US for salaried employees), perhaps they bring in help from other teams, and astonishingly often they just eat the failure.
Under the Moral Mazes view of things, this is because no one can be blamed in a meaningful way. Hiring just enough people to do the routine work is efficient under business logic, and so the people who do hiring budgets are blameless. No one predicted the spontaneous 500 pages of compliance that week, so the project managers are blameless. As a practical matter no one can be expected to quintuple their output overnight so even the employee is blameless; they get punished not for failing to do the work fast enough but for being lowest in hierarchy and therefore most expendable.
Cycling back to the top of the comment, I still think the optionality model is good. It seems like it should be chosen strategically in some cases, like some startups or any business with a lot of variance over the short term.
I usually have so many meetings that I would need to be the 10x programmer to keep two jobs, and it would be completely impossible to keep three jobs. So, for the employers in my bubble, this problem is already solved.
A tool like ChatGPT could be very helpful if it could read the entire documentation in the company and answer the questions about it. Something like search, except you do not need to remember the exact keywords, and it can search across all the systems the company uses (Sharepoint, Confluence, random Word and Excel and Powerpoint documents...). If it could look at the date when the document was modified, and conclude that if the same information is duplicated in many places, the one with most recent date of modification is probably the most correct one. Asking it to listen to the meeting and give me the summary would probably be more useful then reading the official summary.
Put it into ElasticSearch index and give GPT-4 simple query API that it can use by adding some prefix and predefined set of parameters or a JSON so the script would run it instead of communicating this back to the user and give an answer as user response with also predefined prefix. Then it should be able to get questions, search for info, and respond. Worked like a charm for a product database in PoC so should work for documentation.
Yeah, you can do that. How many characters is all that text, combined? Up to ~100k, it just fits in the context window. If it’s more than that, have it describe the contents of each file; then give it the question and the list of descriptions, ask it which files to put into context, and then have it answer from that.