This would be true if it were a zero sum game… But it’s not. Companies reducing their costs through AI a) increase their profits and/or b) can reduce their prices. Scenario a) means the shareholders see the value of their assets go up (which they can sell to buy more goods and services), and/or take larger salaries or dividends (which they’d spend somewhere in the economy). Scenario b) means that consumers see their purchasing power goes up. Yes, laid off worker earn less, but have to spend less for the same goods and services they were previously buying.
In practice what does it mean? Your 401k goes up, goods and services cost less, and more people can now afford luxury goods and services (maids, private coaches, etc.) which in turns creates new jobs where the laid off workers get absorbed.
We’re recruiting AI PMs/builders for almost half of the top 20 venture/growth firms in the US. Not only do they *have* an idea, but many are aggressively retooling for that future. And yet those same venture firms are still hiring associates, why? Diligence will be faster, for sure, but you still need humans to surface stealthy startups and convince founders to take a meeting. AI-augmented VCs become more signal chasing and access/sales-oriented than modeling or market research-focused, sure. But GPs won’t start attending tech parties with 20 year olds.
5 of the 10 best multistage tech investment firms in the world hiring data scientists and AI Product Leads in SF and NYC. 1-6 yrs data exp. required. Must be AI native and heavy Claude/ChatGPT/Python users. Open minded re: everything else. DM me or anyone on my team for details.
I get the shock. The demos feel unreal because the interface between imagination and execution just collapsed.
But here’s the correction most people are missing:
What broke wasn’t physics.
What broke was latency.
We didn’t compress 200 years of discovery into six months. We collapsed the search cost of moving through what was already physically admissible.
Protein folding, materials discovery, circuit design, synthesis pathways. These were never forbidden. They were combinatorially inaccessible. AI didn’t invent new laws. It learned to traverse the space between laws orders of magnitude faster than humans can.
That distinction matters.
Because it means we didn’t cross into “reality is negotiable.”
We crossed into constraint-navigation at machine speed.
Recursive intelligence isn’t magic. It’s feedback loops that shorten the distance between hypothesis, simulation, and instantiation until the loop time approaches zero. When iteration outruns intuition, humans experience it as a discontinuity.
The real danger isn’t unpredictability.
It’s misattribution.
People think intelligence “broke free.”
What actually happened is that coordination failed.
Institutions move at policy speed.
Markets move at narrative speed.
Models move at gradient speed.
That phase mismatch is what feels apocalyptic.
Second-order effects aren’t unknowable. They’re just no longer legible to systems built for quarterly planning and linear causality. The problem isn’t that there’s no roadmap. It’s that we’re still trying to read maps instead of learning to steer.
This isn’t the end of constraints.
It’s the end of pretending humans were ever the bottleneck.
What happens next week won’t be chaos.
It’ll be amplification.
And the question isn’t “are we ready?”
It’s “who learns to cohere before the noise wins?”
Before you all disappear for the Holidays (is it too late already?), here are V associate-level roles we are currently looking to fill in SF and LA, for both emerging and very established firms:
1) Deep Tech* VC Associate, SF. Q1 or Q2 start.
CS or eng. degree + 1-4 yrs of MBB diligence consulting, Google/Meta/Uber APM/RPM, Microsoft's Aspire, LinkedIn's APB, Forward Deployed Engineers at OpenAI/Anthropic/Cohere/Scale AI, etc. or similar structured tech programs.
2) AI Infra VC Associate, SF. Q1 start.
CS or engineering degree + 1-5 yrs Product/Eng. (same as above).
3) Physical AI + AI Infra VC Associate, SF. Q1 start.
CS, eng, math or equivalent degree + 1-3 yrs Product/Eng. (same as above)
4) Enterprise AI (infra + apps) VC Associate, SF. Q1 start.
CS, or eng degree + 1-3 yrs eng. or product at large tech company or startup.
5) Deep Tech* VC + early Growth Equity Associate, Palo Alto. Q1 or Q2 start.
18+ months as a Tech Investment Banking Analyst.
DM me for details.
50%+ of VC dollars are going to AI companies. But, as folks like @alexrkonrad have noted, AI tools are also rapidly changing how VC firms operate. We couldn’t find a comprehensive resource here, so we decided to write it. 📚
I’m excited to announce the launch of one of my favorite @metrics_co content pieces to-date: a white paper on how AI is transforming sourcing, diligence, portfolio management, and LP relations.
We interviewed a lot of VCs (as well as companies building AI-driven, VC-focused tooling) for quotes, case studies, and Q&As across the piece. A fun one was chatting with @WillC_5, co-founder/GP at Riot Ventures, about the intersection of AI and diligence at Riot.
Will regularly invests in highly-technical companies and a lot of his day-to-day consists of sourcing and diligence conversations with nuclear physicists, microbiologists, and robotics engineers. AI, explained Will, “has dramatically accelerated the speed of what we already do for diligence,” and has helped the firm check if underlying technological assumptions of potential investments are sound via “PhD-quality” summarization.
I think it’s a must-read for any VCs trying to figure out where AI might fit into their diligence process. Read it (and a lot more) at the link in comments. 👇
1/
Tech time machine 2020 → 2030
2/
2020: Roll up 12 DTC brands. Get cooked by exploding Facebook ad costs.
2025: Roll up 12 accounting firms. Replace 90% of the staff with AI agents. Lose all your clients.
2030: Roll up 12 humanoid robot startups. Go bankrupt, but monetize the assets by running underground robot cage fights.
3/
2020: Spend 2,000 ETH on a Bored Ape.
2025: Spend $7T on data centers.
2030: Buy the Pacific Ocean for underwater compute cooling.
4/
2020: Put carbon offset credits on a blockchain & build DeFi games around them.
2025: Burn coal to power data centers so China doesn’t win the AI race.
2030: Raise a $100b Climate Adaptation Fund and build 2-inch seawalls around US coastal cities.
5/
2020: No weapons, no fossil fuels, here’s our ESG page.
2025: Missiles defend democracy, so that's impact investing.
2030: Swarm attack systems are peace-tech because wars end faster when both sides run out of humans.
6/
2020: You don't understand the scale, crypto is just getting started.
2025: You don't understand the scale, AI is just getting started.
2030: You don't understand the scale, killerbots are just getting started.
Recruiting up to 6 investors from associate to principal level for brand name VC funds in SF. Focus on Seed & Series A Enterprise AI. Preference for CS degrees but open to wide variety of backgrounds. DM me!
@exec_sum They’re having an insanely hard time finding Wall Street bankers who happen to have 20+ hours available on top of their 16-hour day jobs. What is actually happening is adverse selection as well as hiring bankers outside of the US where $150 goes a much longer way.
Recruiting 2 Principals in SF for high profile Deep Tech VC
#1 focused on space, robotics, nuclear, defense. Great backgrounds: hardware engineer @ SpaceX, Tesla, Anduril, SparkCognition, NVIDIA, etc.
#2 focused on comp bio. AI x life sciences ops/BD background preferred
DM me
Who in my network invests in the creator economy?
Friend looking for seed funding for AI startup that bridges gap between brands, creators, and their audience. Superstar founders worked as leaders at Meta, Google, and YouTube.
PM me for deck or intro.
Recruiting for investment associates & principals for a few amazing biotech-focused VC's + crossover funds in SF, NYC, LA. Open to wide variety of backgrounds. Spectacular academic achievements required. Bonus points for computational bio ML experience. PM me
@graceclarke It sometimes feels like eating broccoli: not always tasty but it « feels » good and I’ll thank myself later for it. Reality is a lot has to do w/ the kids themselves: I have a loving princess that delights us at ballet recitals… and a boy on the spectrum keeping me on my toes
@JoeBiden Naming after oneself an economics model inherited from other people can be a good marketing strategy in good times, but 1) screams hubris and 2) will probably backfire when the economy enters a downturn.