“AI is going to kill eight million driving jobs—then give birth to ten million roles we haven’t named yet.”
Eight million funerals, ten million births. One technology. That’s Jim Bianco’s prediction… and history suggests he’s right.
That paradox is exactly why venture capital is has gone all-in on AI.
The Illusion of the Single Outcome for AI
Virtually every media headline sells the same apocalyptic picture: trucks without drivers, clerks without ledgers, radiologists replaced by algorithms that never blink. But look closer at history and you’ll find the losses are only Act I.
VisiCalc deleted 70 percent of accounting-clerk positions in the 1980s, and cleared the runway for a brand-new species called the financial analyst.
Wages soared, spreadsheets multiplied, and Wall Street wrote thank-you notes to Silicon Valley.
Bianco’s forecast for autonomous vehicles follows the same blueprint:
Erase the driver.
Collapse the cost of movement to near-zero.
Watch entrepreneurs sprint into the void.
No garage? No driveway? No problem.
Imagine the businesses that bloom when a twenty-minute delivery window costs a couple bucks and a mobile clinic can park at any curb.
The Moment Everything Gets Cheap
If displacement is inevitable, so is decentralisation. Right now the gatekeepers are trillion-dollar platforms and GPU farms the size of townships. But the price of artifical intelligence is falling the way storage and bandwidth once did… fast, then faster.
January’s buzz around DeepSeek wasn’t about a clever model. It was a proof-of-concept that you don’t need a nuclear-powered data centre to train something useful.
Once the toolset crashes through that cost floor, the real sprint begins, as it did when building a website went from “hire AOL” to “sign up on Blogger”.
Three founders, one credit card, a model fine-tuned on their own data; that’s the point on the curve where massive incumbents stop buying talent just to bury it, and start losing sleep because they can’t bury everyone.
Four Fault Lines Ripe for Startups in AI
* AI Toolchains for Mere Mortals
Low-code platforms that let a manufacturing COO fine-tune a vision model without hiring PhD #43.
* Edge Intelligence
Chips and cooling that move inference from hyperscale warehouses to forklifts, drones, and irrigation pumps, cutting latency and cloud bills in half.
* Data Franchises
Proprietary corpuses… think anonymized medical imaging or soil micro-maps that feed narrow models rivals can’t clone.
* Interface Alchemy
Products that make AI invisible. The killer app won’t wave “powered by GPT-X” banners; it’ll just fix supply chains at 3 a.m. while you sleep.
Those plays share one trait: they get stronger (not weaker) each time the cost of cognition drops another notch.
Why Venture Capital Wins AI
Capital doesn’t fear chaos; it profits from it. When a technology swings from monopoly to free-for-all, two things explode: the number of bets you can place, and the asymmetry of their outcomes.
But timing is cruel. Back Cisco in 1999 and you’re still waiting (to this day) for that trillion dollar market cap, despite everyone predicting it would happen before the turn of last century. Back Google in 2004 and you outrun the index for a generation. The difference? Decentralisation had finally arrived; code was cheap, browsers were everywhere, and dorm rooms were the new R&D labs.
AI’s cost curve is bending in exactly the same direction.
What Founders Should Hear in This
If you’re a startup: don’t waste breath promising to “disrupt” the mega-caps on their own turf. Instead, look for the second-order life-styles and supply chains their platforms cannot see. When drivers vanish, how does retail layout change? When edge models run on AA batteries, which rural industries reboot?
If you’re a VC: remember that scale is a moat, until the drawbridge rots. The next Amazon of AI will probably start as a feature nobody in Mountain View deems worthy of a product briefing.
I left my conversation with Jim Bianco feeling optimistic. Yes, eight million jobs will likely evaporate, just like so many ledger books. But venture capital isn’t in the obituary business; it writes prologues.
Aaron
