SWL Week in Review - Aaron Levie, Voight-Kampff, Retail Fomo
More or Less with Aaron Levie… on AI, Crypto, and Politics… Aaron is one of the great internet shit-posters and I had particular fun having him on MoL — he is great to bullshit with. This one is actually the rare podcast worth a 1x speed listen — The big debate really is if AI is massively deflationary / crushes the value of software businesses (my bet) or a huge propellant. We will see!
Some Other Hot Takes:
- Bitcoin 100K - the average of zero and infinity :: I am really glad I showed up at a dialog session in March 2013 that finally tipped me into racing home, going to western union, and sending money to bitstamp to buy BTC…. Lots of people are using this moment to talk about wealth generated from long-tail-low-conviction bets, narrative shifts, blah blah blah… but here is the most simplistic way to think about things - the mean of zero and infinity, is infinity…. And very very few things in life have theoretically ‘infinite’ upside — when you find one that big, you write the check.
- Don’t Miss the ‘Swerve’ - The Emotional Reality of Retail 1000x Opportunities :: speaking of which, one thing I think is particularly consternating to many folks about Bitcoin is that it actually has been a retail opportunity for ages… as ‘retail’ has access to things like BTC, options trading, etc. I think there is a real shift in mentality and sense of FOMO investing — the average person thinks less about the 5% vs. 7% return, and more about covering tails / not missing the ‘swerve’
- Believing Harder vs. Believing Differently are different forms of investing — I am a believe differently guy, but believing harder tends to be a better business model
- Granola kinda illegal, definitely awesome — the more I use granola the more I like it as an investor — if you are a VC the meeting notes it pulls out are very well tuned, excellent… and the UX is great — there is that nettlesome problem of legality…
- Can You Be an RIA and be good at seed investing? Will Quist my partner made a pretty compelling and interesting point that no, you cannot — … many make the counter point of Andreessen… I would snakily question whether he is a good seed investor anymore… asset manager yes, but seed investor? … though I agree his speaking seems relatively unconstrained by compliance so maybe you just need a very large legal team to be both an RIA and a good seed investor…
- I Think I Need an 8 Week ‘Slow Intern’ Program — something I am really considering for 2025… rolling 8 week pool-house internships / crash course in VC for you, new ideas, networks, access to some crazy kids for me? …. yes, we do a 24-36 month associate thing, but maybe a rapid fire young person thing as well?
- The Non-Agentic Harris Campaign — great encapsulation via Nate Silver via Nellie Bowles — this really caputures how I feel at this moment — say what you will, but Trump and these Czars are… Agentic — and we do need leaders who take action vs. NPCs.
- 10% of global elites went to Harvard? — wild study which explains why the pressure on Harvard is so intense / why people care and why it matters — its share of the ‘elite’ is pretty wild … makes the endowment (and pressures) on the university seem if anything under-indexed…. So yah, it is worth some focus.
Happy Weekend,
Sam
P.S. congrats Singerman on one of the more epic runs in the history of venture capital
P.P.S. - Voight-Kampff companies… an AI investment theme I actually believe in as the economics of scams / fakes goes from alignment (scammers don’t want to be too smart), to a war of attrition.
In Blade Runner the AIs get so good it is impossible to tell who is a real human and who is not — the only way is the Voight-Kampff test.
We aren’t there yet in the real world (and don’t hold your breath) — but we do already hugely have this problem all across the internet with synthetic content — AI resumes and cover letters delivered at scale — personalized AI driven drip sales campaigns, etc….
We have always had these problems — but as the cost of content / generating believable but fake content goes to zero, what is interesting is that the strategy / economics of ‘fake’ changes in a way that makes the problem way way more acute.
- HISTORICALLY IT WAS RATIONAL TO MAKE FAKES EASY TO DETECT :: When ‘fake’ is relatively expensive to run (think Nigerian prince scam emails) — the economic strategy of ‘fake’ is actually to make it clear to anyone who was engaged that it was a ‘scam’ and just catch the tail end people that are dumb enough to fall for it (or really need a friend so fall for it on purpose)…. This is annoying of course, but means that actually real people and scammers are aligned to make it easy to tell what is fake and what is not
- NOW THE ECONOMICS FAVOR REAL FAKES / REAL TREACHERY :: But as ‘fake’ becomes nearly free and ubiquitous and personalized, then all of a sudden the ‘economics’ of fake content shift, and it makes way way more reasonable and strategic to actually try to trick everyone not just really dumb or really lazy people… This is a massive step-up in the practical cost of ‘fakes’ for everyone… all of a sudden the ‘fakers’ and the targets are no longer aligned that it should be easy to tell what is fake and what is not
Enter Voight-Kampff companies — that means that all of a sudden there is a REAL NEED for platforms and systems to validate the legitimacy of content / resumes / sales pitches, etc… reality becomes TRUE warfare…
I have already started investing in (and building) some companies and products around this theme — but it took a sauna with Mike Galpert yesterday to realize that this is an important ‘theme’ not just a few one-offs…
I am short investing in the AI companies, but long on the companies that orient around identifying and rooting out the fakes.
sam lessin | @lessin | want to talk | podcast | stop | to get added to the mailing list email sl@wlessin.com
Similar newsletters
There are other similar shared emails that you might be interested in: