Matthew Lewis

we're hiring a killer operator. (lightly) technical & creative ;) - turn processes into smooth machines - like being the bridge with engineering - detail-obsessed & lover of data - great at getting people to do things they have to do - treat everything as figureoutable - love design/creativity - hater of AI slop typical suspects include: PMs, technical PMs, banking, consulting, project management. looking for full-time contractors that can start now (remote, anywhere), or full-time roles in SF. I can promise that: it'll be hard, and you'll have a lot of fun ;) DM me if you're interested!

Thinking of organizing a group hike along Dipsea Trail next weekend. If you're in SF and interested, shoot me a DM/text!

@wquist @villi Will continue to shamelessly tail you in support every time I see you say it

@villi Private. Credit.

LPs asked me what could derail the current boom in tech and AI. My answer was I don't know. But, when everybody is levered to the gills, every asset is priced to perfection, and incompetence and corruption are running wild, you just don't know when/where it may come from.

i <3 operating leverage

.@NYCMayor Mr. Mamdani, this would be an excellent feature to implement for the @nycmarathon. I’m sure we can figure out how to safely make it happen by November.

@FinnMurphy12 @FinnMurphy12, you're welcome to join @adlabossiere and me for our recurring meeting "The Quest to Find the Best Guinness in New York" which just results in us meeting at Swift every other Tuesday

best Irish pubs for a St Patrick’s day Guinness in nyc 1. Swift Hibernian Lounge 2. Four Faced Liar 3. The Long Hall (nyc) 4. The Craic 5. Iona (actually a Scottish bar but a good pour) 6. Dead Rabbit Be sure to celebrate properly and have at least 2 pints in each today ☘️

@owl_posting So excited to read this

Reasons to be pessimistic (and optimistic) on the future of biosecurity owlposting.com/p/reasons-to-b… "It was such a fun read (if you can say that about an article on weapons)!" —a glowing review from an early reader this is (once again) the longest article I have ever published at 13,000 words. it involves interviews with 16+ researchers/VC's/policy folks in this field, and discusses basically every single facet of biosecurity that i could find. topics include: how machine-learning in rapid response therapeutic design may work, the financial status of the customer base of biosecurity startups, why agroterrorism feels extremely likely to me, and a lot more i admittedly started the essay pessimistic that this subject matters at all, and i end it surprised that it doesn't keep more people awake at night. im not a doomer about it all, but i can see how people become one. very grateful to the people who decide to spend their career (or some fraction of it) working here, and especially grateful to the ones who helped teach me about the subject

@tdozzi Thanks Tom!

Great piece by Matthew here

@wquist Always was x.com/mp_lew/status/…

Is this the plot? Private credit crunch creates overall illiquidity and compute build out slows. AI startups can't grow and / or the math really stops working. Than rate cuts into or out of the midterms. AI heats up again, but then power is the issue and AI scaling becomes a long slog...

@Bfaviero 👀

“The Taste Company of San Francisco” could probably raised $100m

@Trace_Cohen good point @grok who do you think is being portrayed in the original post of this thread? make no mistakes

@mp_lew If they got carry you need to be at a fund at least 10yrs so that narrows it down a lot.

Reveal yourself

When you hire a new employee, you onboard them, drip them context about work/best practices over time, and review their output before shipping. The company dictates the employee's work, not the other way around. With reps, they can earn more autonomy. It's naive to assume agents will be fully autonomous "workers" with little visibility and oversight. Ramp Agent Cards is the first product appropriately bridging human expectations and norms with agent leverage. I expect more to come in short order.

@almostcmb matthewlewis.xyz/blog/2024/07/0…

This is mostly a matter of semantics. "Deep tech" itself is a lazy term. I expect we will stop seeing it before long. Reminds me of how funds used to call themselves "internet" investors. It's also a relative term. Cloud infrastructure used to be deep tech. Where do you, as an investor in 2026, make the judgment call to draw the line? I understand most investors not feeling comfortable underwriting a quantum company's ability to reach 10,000 qubits. But we are in a very specific technology cycle and macroeconomic circumstance. Engineering and science risks are so deeply interwoven into the overarching story that eschewing them as an investor lops off a fair amount of opportunity. As an anecdotal example, it's nearly impossible to have a grasp on the AI arc without being able to articulate the thesis behind model specific silicon and photonics. Even if you aren't interested in underwriting materials science, there is still huge justification for immersing yourself in the most important R&D efforts of today. Global supply chains and distribution strategies are evolving around these efforts. Component part manufacturers in wire harnesses, composites, actuators, and motion systems are standing up to serve new-age OEMs. Robotics companies are partnering with financial sponsors to distribute into industrials customers. Are you experienced in understanding supply chain bottlenecks? I hope so if you're professionally allocating capital. But before you have a chance to partner with or underwrite any of those opportunities, you better understand the why now...and the BOM cost of a humanoid...which you can only get to if you give yourself the freedom to explore... Also, the hardest problems with the biggest prizes tend to attract the best talent. If you're in the business of early stage investing, you definitionally care about finding and backing exceptional people. A natural response to this might be "Hey, my network has never been physics labs, so I am not missing out on anything." That is likely wrong. There are two founders in our portfolio executing exceptionally well on hard tech businesses (anti-drone defense and robotics) that spent the decade prior building in consumer, which is where we first got to know them. "You can just do things" etc etc. I agree with most of Sam's core premise...competitive markets warrant specialization and questions around right-to-win...but also, no advice in the world has killed more careers (or alpha) than "stay in your lane..."

@nikhilvnamburi Correct, but they're not venture theses. Illiquidity for 10 years is the whole point.

many popular VC investment theses can be more cleanly expressed via public stocks/lab equity

On the social perception of AI... I think we'll see at-scale demonstrations for agents' rights by the end of 2027

Early-stage venture has always been -- in the most financial sense -- an exercise in exploiting imperfect information and arbitrage. That hasn't changed despite the current "hot market." Gotta go and find the next thing. Don't lose the plot.