24) Value Compression

Behind-the-scenes building Vambrace AI, a company on a mission to figure out its mission. Please pardon the stream-of-consciousness style. Subscribe to follow along or visit the site here:

(typos are to make sure you’re paying attention)

Introductory Remarks

Dear Vambracers —

In last week’s post, Customer Success, we explored the recipe for a successful customer engagement and mapped out some guiding tenets as I work to build a customer-focused enterprise. The importance of satisfying your customers and delivering on what you say you’re going to do is, I think, one of those super fundamental and obvious things that is so obvious that it’s easy to overlook. So I don’t want to lose sight of its importance as I start to push towards delivering some real outcomes.

Value compression

In today’s post, I want explore the current state of the labor market and discuss some interesting dynamics I’m starting to encounter around the perceived value of labor (and its forthcoming compression).

Value of labor

We have a system that assigns some economic value to labor performed over some fixed amount of time. The economic value that we assign to a unit of labor has, theoretically, been dictated by market forces over the past ~60+ years and is broadly co-incidental with (and related to) the rise of the white collar class (and the MBA-alization of business culture).

As I think about pricing, I don’t have a good system for understanding the fair market value of my work. I could try to look at what market-rate is by legacy consulting firms, and then try to roughly map my output to their output, and then adjust for lack of firm reputation and stuff and go from there. Or I could just assign some arbitrary $ / hour to my time and then roughly go from there. But if I tried to do that I’m not sure that I could root my $ / hour in anything other than vibes.

Not to mention how quickly I can move relative to legacy peers with the power of AI. So the value of my time is maybe higher because I can do more with less, and so a credible argument to made that I should charge a premium for my time but then I’d spent less time in the aggregate on certain things that would take legacy firms some amount of time, and so I’d charge less for the same work and finish it quicker.

And I think that’s where things get tricky, because the efficiencies made possible by AI are not marginal—they’re massive. As an example, I was able to rebuild my first customer’s website in about a week. I didn’t do anything too sophisticated but I gave it a complete face-lift over the course of hours. Similar work might have taken a legacy dev shop weeks if not months—and I’ve gone through processes in the past where there’s been repeated back and forth and it’s taken a long time to get things just right—and it was difficult and grueling. And that’s just not how the world works anymore. So do I charge what the dev shop might’ve charged (maybe $10K+?), or do I take like $100 / hour (or some arbitrary number) and multiply it by like 8 (and even that’s kind of generous)? I really don’t know.

Perceived-value-of-labor arbitrage

This brings us to the “perceived value of labor arbitrage.” Some fixed output, say, a website, would have taken [x] hours and cost $[x] * [y] / hour in the past. Now the time it takes to complete that same fixed output has radically shortened and the $ / hour realistically has also compressed—because really if you take humans out of the loop entirely it’s just some $ / token equation. Of course humans have to prompt and review and guide, and there’s real economic value there, but I just can’t overstate how big the gap is that I feel forming between legacy prices and modern real-cost in an AI-first world with an AI-first approach.

So then the question becomes how much of that labor arbitrage should I take for myself and how much should I “give back” to customers? And I’m sure people everywhere are grappling with this very thing.

Competitive dynamics of value compression

In my case, I’ve decided to try to give a decent amount of the value back to the customer and bake speed, cost, and quality into my value proposition. I’m trying to position as someone that can go deeper for cheaper, and more quickly than others. I don’t think that the efficiency gains made possible by AI should be gate-kept from customers—and I want to evangelize the power of the tools while also delivering for customers.

I think this also serves to reinforce a customer-centric and trust-oriented relationship with people. Part of my push into entrepreneurship in general is that I’m sick of performative work, and so I don’t want to obfuscate how quickly I can do things.

I also fundamentally subscribe to the belief that the real promise of AI lies in our ability to pursue bigger and more challenging problems in tandem with AI—on a societal level. And so as I think about my personal and professional development, I want to leverage AI to become faster, more efficient, more effective, and to think more deeply about things. And this is the market position I want to take as I try to win more business—and as I seek to conscientiously attract like-minded customers that want to move quickly, do more, and outcompete peers.

Carveout: Psychological impact of value compression

Something I’ve personally been contending with is intense and acute imposter syndrome borne out of how much I’m able to do with AI (which maybe sounds like a humble brag but I assure you it is not). I grew up (professionally) in an environment that put a premium on pain and struggle, and I put those things on a pedestal as critical and necessary components of any worthwhile work-product. And I’m still struggling and striving and learning and pushing, but there’s also a lot less acute struggle. And then I can also move so much quicker and do so much more so much faster than in the past. It’s all very disorienting and intoxicating at the same time. I can’t tell if I’m just drunk on the liquor of LLMs, or if I’ve discovered a new frontier of human performance and am experiencing natural aimlessness associated with uncharted territory. Am I a courageous cartographer or a delusional drunk blubbering on a barstool in the AI Alehouse? I don’t know. And I’m not sure when I’ll find out. And that’s just not fun! But, alas, c’est la vie, mon ami.

Looking Forward

An economic reckoning is on the horizon for the broader professional services industry and I’m not sure where things will go. I could see massive organizational restructurings for legacy consulting firms and/or I could see all companies effectively in-sourcing “consulting-type” tasks. Because AI democratizes many tools that previously required specialized labor and expertise—and now you can get 90% of the way there on your own. So why spend hundreds of thousands of dollars on this stuff?

Sincerely,

Luke