Summary: AI-accelerated teams of the world’s very best programmers will be tiny but create about $100M/person in startup company value. How can UX keep up with these SuperTeams? Is SuperUX good enough? UX needs its own restructuring for efficiency.
We’ve known for 40 years that good programmers are around 20x as good as bad programmers. The very best are about 100x. Most companies can forget about hiring a 100x programmer, but if you’re a super-promising startup, those are the people to hire for fast go-to-market.
When further accelerated by AI, the 100x programmers will not pull further away from normal programmers because AI tends to narrow skill gaps. An AI-augmented 100x programmer may only be 50x compared to the new performance of an AI-augmented normal programmer. But the AI-augmented used-to-be-100x will be 300x the old norm (Throughout this article, I’m using multiples of the 2022 average level as my performance indicators.) And the good-but-not-super geek who was a 10x programmer in 2022, will be a 50x programmer in 2027 when augmented with AI, again relative to the old norm.
This means that startups that score those very top programmers (in the top 0.01%, so only one in ten thousand programmers) can achieve unhuman feats with AI augmentation and build major products with a team of 10 super-programmers. An interesting video from Y Combinator (Silicon Valley’s leading startup incubator) discusses the potential of such super-startups where 10 super-programmers build a billion-dollar “unicorn” company.
There are about 3,000 100x programmers worldwide, or enough to staff 300 unicorn startups of all-SuperGeeks.
Sadly, SuperGeeks are still geeks. They don’t know how normal humans think. For their product to be good, the SuperGeeks need UX help. As I argued in my piece Peak Programmer ≠ Peak UX, there will be a substantial need for UX jobs in the future.
SuperGeek vs. SuperUX: can the two work together and build super products? (Ideogram)
Is SuperUX Staff Good Enough?
The world has about 3,000 SuperGeeks, who are 100x programmers without AI and will be able to perform at about a 300x level with the improved AI copilots they’ll have available in a few years.
How many SuperUX does the world have?
There are about 3 M UX professionals in the world. Thus, the top 0.01% constitute 300 SuperUX staff, or one for each of the potential 300 SuperStartups that could be created from those 3,000 SuperGeeks.
But are we good enough to do this job? I think it’ll be tough for us to measure up to the SuperGeeks.
Two reasons:
Performance variability seems smaller in the UX field than among programmers. Of course, the IQ distribution is a fact of nature and dictates that some are better than others in any group of people. Still, while UX excellence obviously depends on IQ, it seems to track the underlying distribution less precisely, especially on the higher end. Thus, while the very top geeks are around 100x better than average geeks, the very best UX designers may only be 30x better than average UX designers.
AI acceleration is lower for UX (and most other knowledge workers) than for programmers. While a SuperGeek may go from 100x to 300x with AI assistance, a SuperUX may only go from 30x to 60x with full AI support. (An average programmer may go from 1x without AI to 10x with future AI, whereas an average UXer may go from 1x to 3x when getting the better AI we expect in the near future.)
In other words, SuperGeeks will be 300x talents, whereas SuperUXers will only be 60x talents. A factor 5 difference in elitism between Geek and UX.
A second issue is that UX comprises many subdisciplines, most importantly research and design, and that any one UXer is only truly great at a subset of the field. For average UX work, this is no problem, since AI is a skill accelerant and allows people to perform well enough in skills beyond their core strength. For most jobs, most UX professionals will become “UX Unicorns” able to do it all.
However, if we want superior excellence in UX performance rather than merely good enough, UX Unicorns do not cut the mustard. To match the performance of the SuperGeeks — and more important, to earn their respect as equal partners — our SuperUX staff would need to specialize, meaning that the team would require at least two of them: a designer and a researcher. This goes against the concept of “pizza teams” that are kept tiny to minimize communications overhead so that they execute at lightning speed. (The term derives from the ability to feed the entire team with two pizzas.)
A pizza team of nerds must be kept at a minimal size to ensure extreme efficiency and high shipping speed. Every extra team member slows down the team, which counts against UX if we need to add two warm bodies to do our job. (Leonardo)
I’m arrogant enough to believe that I might have been one of those SuperUX people 30 years ago. (Not now: I’m too old to work the 70-hour weeks required for superior performance and keeping up with the SuperGeeks.) Despite this claim, I know I could not have been the only UXer on a pizza team to ship at light speed. I could have done all the research at Warp 9, but my designs would have been subpar. (I can’t even design a decent color palette — though this specific task is indeed one that AI will do perfectly in a few years.)
User research at Warp Speed: I could have done that 30 years ago. (Now, I’m too old and lazy.) But I would have had to be teamed up with a SuperDesigner to produce the UX deliverables needed to feed a pizza team of SuperGeeks. (Leonardo)
Resolving the SuperUX Paradox
SuperGeeks need UX to deliver the products that will make a 10-person pizza team worth $1B. But the SuperUXers are probably not good enough to earn a place on the pizza team. How can this paradox be resolved?
If you’ve read my past writings, you know that I’m usually quite sure of myself in telling you how to solve problems. But I’m honestly stumped at predicting how this one will play out. I see 3 possible solutions, and the future will likely be a blend of them all, though I don’t know the relative contribution of the 3.
Solution 1: Super-SuperUX
While SuperUXers are likely not good enough to be on equal terms with SuperGeeks, this may not be true for all of us. The normal distribution of IQ and other talent drivers is still at work here. The very top of SuperUX is likely good enough. Not 300 UXers in the world, but maybe 100 could be in this super-duper category. Not enough to serve all the SuperGeek teams, but some SuperGeeks could be joined by a single Super-SuperUX.
These Super-SuperUX talents might be good enough to have sufficient Unicorn powers beyond their core competency. For example, I know a very good designer who is only decent at user research. Clearly not good enough to be the sole research source for a world-class product, even though she could handle the design part of the assignment. But that’s now. In 5 years, with AI augmentation? Maybe top talents like this become SuperUnicorns, and even their weaker skills grow strong enough.
In the Legion of Super-Heroes, Triplicate Girl was a valued member (her superpower was the ability to split into 3 people), but other members like Superboy and Saturn Girl were considered more powerful. (I was a nerd as a kid; what can I say.) Similarly, among the hallowed SuperUX ranks, some will be more super than others. (Leonardo)
Solution 2: Two or Three SuperUXers Team Up
I’ve already hinted at a likely solution below that elevated Super-Super tier. The “regular” SuperUXers (still world-class talents, mind you) could pair up (a SuperDesigner with a SuperResearcher) or maybe form a small team of three (visual design, interaction design, user research).
Combining people with different skill sets solves the multidisciplinary problem of UX. But it undermines the pizza team ethos and makes the entire company less efficient. Maybe the SuperUXers have to accept less stock options than the SuperGeeks because of the damage they do to company performance.
Clearly, this is a suboptimal solution, but it will guarantee excellent results on the UX side of product development, even as it undermines the speed of the project at large. This is a likely compromise in many cases.
Solution 3: Outsource UX to StrongUX Consultants
The third option goes against everything I believe, and yet I suspect that it will become the most common solution: don’t integrate UX with the core development team, but outsource it. Engage a UX consultancy. (Usually, it’s better to invest in building UX skills and capacity inhouse than to pay another company to upskill their consulting team while they try to figure out how your industry works.)
There are endless problems with keeping UX separate from the product team, but there are two huge advantages:
By outsourcing, we can utilize a lower level of talent, say top 0.3%, instead of the 0.01% required to be an equal member of a SuperTeam. They are paid a flat fee instead of company stock, and they don’t participate on the core team, where they would slow daily work by filling the room with too many warm bodies. This allows 30x as many UX people to join the game, or about 10,000 people worldwide.
Because they are segregated in a separate company, the outsourced UX team can now include many members, each with strong skills in specific subdisciplines. Presto, UX-Unicorn problem solved.
Keeping programmers and UX designers/researchers in separate companies allows both teams to be small and efficient (good). The likely deal will allocate 99% of the value to the programmers (bad for the UX folks). (Leonardo)
In my article Commodification and Pancaking of UX, I pointed out that UX will mainly be spread thin and that there will not be as much of a career future in climbing the old-school greasy pole of corporate management advancement. However, I did predict strong career prospects for UX experts. This type of “StrongUX” consultant is likely to be one such lucrative future for UX talent who’s not in the very top 0.01% but still very strong.
I previously predicted that UX consulting would change from being shaped like a dromedary (one-humped Arabian camel) with most work in the middle to becoming shaped like a Mongolian camel with two humps. One hump will be comprised of scalable firms providing low-end UX design to the masses. The other hump is likely to be several thousand small (5–10 consultants) boutique firms, mainly drawing from that pool of the world’s top 10,000 UX talents.
Current UX consulting (that dromedary shape) mostly targets enterprise clients. As long as your consultants are in the top 5% of the talent pool, they will be smarter than most of the client staff, thus able to charge exorbitant fees. And even though the smartest people on the client side will be better than the UX consultants, they can be soothed by sleek-looking slides and smooth presentations.
Most current UX consultants won’t be able to transfer to those StrongUX boutique firms because they’re in the top 5%, not the top 0.3%. Such good, but not great, consultants are advised to jump ship and get corporate jobs while they can. Membership in that top 5% will secure a cushy enterprise career.
However, as mentioned, there are probably 10,000 UXers who are good enough to hold their own as the StrongUX outsourcing partners for a SuperGeek company. This may not get you $100M/person in stock options, but it should pay extremely well.
Tier-2 Strong Talent Forecast
So far, I have analyzed the potential for a tiny elite of superpowered talents to build valuable products in tiny teams. If a 10-person team builds $1B in value, that’s $100M per person.
What about the next level down in talent? There are about 10,000 UX professionals and 100,000 programmers in the world who are extremely good, even if they are not at that superpower level of talent. I’ll use the terms StrongGeek and StrongUX for this level. Remember that these “Strong-level” talents are the world’s top 0.3% in their field, so they definitely are excellent. They’re just not at the superpower level of the top 0.01% of the talent pool we discussed before.
A StrongGeek probably programs around the level of 20x without AI and 100x with next-generation AI.
Similarly, a StrongUX designer or researcher probably produces designs or research insights at a 10x level without AI and a 25x level with AI.
Whereas the SuperTeam is pizza-sized at about 10 members, a StrongTeam might need 36 StrongGeeks and 4 StrongUX members, for a total of 40 members. That’s a big takeout order for the all-hands meetings. In fact, the very need for all-hands meetings instead of everybody knowing what everybody else is doing through osmosis is an indicator of the communications overhead and performance penalty from scaling up the team size.
One huge advantage of the StrongTeam over the SuperTeam is that it should have about 4 StrongUX members. This means that it has positions for dedicated experts in research, visual design, and interaction design. (The 4th UX position on the team can either be one more designer with a specialization that’s suited for the product at hand or maybe somebody who specializes in a more esoteric form of UX design that’s needed for that product. Maybe sound design or Augmented Reality design.)
Even though the Strong-level professionals are excellent, they won’t perform at the same level as the Super-level pizza teams. So instead of creating $1B value, maybe the results of their efforts are only worth $400 M. That’s $10 M per member, which is a huge payoff for joining the team. And since we’re talking about the world’s top 0.3% of programmers and UX specialists, it’s fair enough that they get well compensated. Still, this is only 10% of the value created per member of the SuperTeams.
Career Advice
What can we conclude from all this? If you are truly the top of the top of UX talent (0.01%), beware of joining a StrongTeam, even though it seems (and is!) excellent. If you qualify, the expected value is 10x higher if you are accepted into a SuperTeam.
However, only the world’s best 300 UX specialists are qualified to be on a SuperTeam.
Even if you’re extremely good, it’s more likely that you are Strong level, not Super level.
A StrongUX candidate has two options:
Join a StongTeam: estimated payoff $10 M.
Join one of those StrongUX boutique consultancies with about 5–10 great UX members who specialize in outsourced UX design. Salaries at such a high-powered firm should be high, but most likely less than $0.5 M per year.
Money-wise, option 1 is clearly the best, but it comes with a bigger risk since not all these companies will succeed. In contrast, if a boutique UX firm is good, this will be known and produce a reasonably steady stream of clients. There are always ups and downs, and some years you won’t get a bonus. But working at a boutique consultancy is much less stressful.
I can’t say what’s best for you — depends on whether you value money and excitement higher or lower than a relaxed lifestyle. Both options are great, as is appropriate for somebody in the top 0.3%.