I started my tech career in the early 90s, and worked as a Perl developer for eToys dot com based in London and Santa Monica during the dot-com boom. I was there when eToys hit 80 bucks a share at around a 6 billion market cap, and was delisted about 18 months later with its assets sold off to KB Kids for $2 million. Every morning I would watch CNBC while getting dressed at corporate housing in Santa Monica watching that share price plummet and plummet and keep going down with no bottom in sight. Turns out there was almost no bottom – at least in the case of eToys.
So when it comes to the dot-com bust, I’m familiar.
These days I run a small (40 people) fast moving tech business in cybersecurity as 50% owner and CTO with my wife who is CEO and owns the other 50% of the biz. We’re big in our space – the biggest in fact, with millions of free customers, and 6 figures in total paying customers.
We continuously strive for engineering excellence and I’m proud to say that my team and I are power users of AI, have built AI products, and are pushing the capabilities of AI to the absolute limit and beyond. Most of our breakthroughs happen in the dark because we’re in cybersec and opsec is a big deal in our space if we’re to stay ahead of the bad guys. But the GitHub CEO’s comment about AI increasing ambition is our north star. We’ve massively increased our ambition, and by extension our accomplishments through extensive use of AI.
So when I read articles like “Fund managers prepare for ‘reckoning’ in US tech sector” in the FT with a quote like “This looks like the dotcom [boom] on steroids.” I can’t help but marvel at the level of naivete and idiocy this exhibits.
The dot-com was all about the pre-revenue and pre-profitability IPO. The AI boom is remarkably different, in that customers like my company can’t pay foundational model providers fast enough. Our current spend across providers is around $150,000 per year, although it varies wildly from month to month as we hop from one leading model to another, based largely on scores like ARG-AGI-2 and others, depending on our needs.
But spend we are, and hand over first is the order of the day. ChatGPT may be the fastest growing product in history, but what is hidden is the speed of revenue growth at many AI startups with compelling products.
It’s tempting to hedge at this point, and say something like “yes there will be big failures” – but one doesn’t need to hedge, because sophisticated investors hedge through diversity in their investments in the sector. So provided we have winners the size that we anticipate, the hedge is built in. They’re investing in 100 failures, banking on one or two of them being the next Google, Microsoft, Oracle, etc.
But I predict the winners will be bigger, and more powerful, because every winner will control the source of half or more of the future of labor, and the most powerful cognitive and problem solving and innovation capability on the planet. I don’t think AI is a labor replacement. It will more than 10X the global labor force. AI and the workers of the world will both win. We’ve already termed a skilled AI practicioner in our org as an “Operator”, which is a blend of dev, ops, a fundamental understanding of AI, and the ability to generate and ship.
Betting against AI isn’t without cost. If you’re paying rent on put options as they decay in time, it has a direct cost. But if you’re simply ‘noping’ out of the AI sector by not holding the relevant stocks, the opportunity cost if this is not a bubble is catastrophic for any fund that plans to be in business in a few years – or any fund manager who plans to have a reputation a few years from now.
Incentives, being the driver of all behavior that they are, are driving journalists to cash in on schadenfreude by churning out “its just like the dot-com bubble” articles for rags like the FT. And lets face it, the FT is a euro-cynical rag these days, cashing in on an economically declining readerships hopes that the other guy will fail. What they’re not taking into account is the cost if they’re wrong – and the cost to their believers if they’re wrong.
So what does that look like? Like this: We have underestimated the amount of revenue and profitability that AI winners will produce. We have underestimated the amount of new value it will create. We have underestimated the number of new jobs AI will create. And we’ve underestimated how much that vast new contributor to global GDP will stimulate spending in other sectors, even sectors we consider ‘legacy’ or dying. For the cynic, this potential future reality is the epitome of terrifying. Its unthinkable.
So the rags stay in business by selling content that keeps the cynic warm at night, and the innovators continue to be so excited about the future, and what we will create for the people of the world – the world that we care so much about – that we don’t need to be kept warm at night because we are not sleeping. We are creating with optimism as fast as we can and we are excited for your future.
Leave a Reply