On Synthetic Cumulative Culture: a concept I have been thinking a lot about lately is that of Cumulative Culture. There are many definitions for it but a simple one would be that it’s our capacity to build on the cultural behaviors of our predecessors, allowing increases in cultural complexity to occur such that many of our cultural artifacts, products and technologies have progressed beyond what a single individual could invent alone. It’s been reported that printing and digital technologies have fast-tracked the accumulation of culture through information aggregation but this information was produced by humans or at least by non-smart ML algorithms designed by humans. So how will this change now that it’s the first time we have agents producing synthetic data and information? We already have LLMs trained and finetuned on synthetic data generating brand-new information (even knowledge) that humans use to advance research/innovation/experiments. Does this mean that it’s the firs time in human history that the accumulation of culture will be a mix of human and synthetic data? I believe the answer is yes and it’s not something we should be afraid of. But in that case there are so many interesting questions that arise. First of all, should we even call it a culture? Will a synthetic cumulative culture bring a solution to the famous problem of “lack of good ideas”? It is possible that a synthetic culture is not only of “higher quality” but can also be spread and diffused faster and easier to more humans enabling faster progress. There is also a scenario where AI fosters the development of a man-animal cumulative culture. We have already seen cases where apes are copying human behavior for their survival but also examples of LLMs trying to decode animal sounds. So a cross-communication between animals and humans isn’t a sci-fi scenario anymore.

On the power of visualization: I haven’t stopped thinking about the recent AlphaFold3 announcement. Of course, it has a profound scientific significance on biology and drug discovery but it’s also a powerful reminder on the importance of mapping and visualization. It’s only when we can understand the structure and mapping of something that we can actually engineer it. Crystallography was already doing wonders in biology but GenAI is apparently the perfect tool for discovering all the hidden patterns, connections and interactions between molecules. That leads to the obvious questions: what else, which is currently a mystery box in regards to its shape/structure, could we “discover” by mapping/visualizing through GenAI? What hidden patterns can AI discover in multi-dimensional data? First candidates: neural networks, social graphs, material science, complexity datasets.

Living Onchain: it’s been almost 2,5 years since I started working full time in crypto and it’s still ironic to me that most professionals in this industry aren’t spending enough time/energy onchain. This post from Fred Wilson struck a chord. I mean, what’s really the point of being in crypto if you aren’t onchain? For an industry obsessed with transparency and accountability, there is way too little skin in the game. Yes, there is tons of activity on DeFi which I guess it’s fine but makes the whole industry seem like a huge casino with a stagnant monoculture. There are so many interesting consumer crypto apps that I literally find more fun/interesting to use/interact with so I will be sharing some of them hoping to encourage more crypto and non-crypto folks to try them out.

• Blogging: My blog will remain on Substack for the near future but I have already started cross-posting my thoughts on Paragraph and t2 world. You can now finally start minting my posts!

• Music: increasingly using Sound.xyz more than Spotify and buying music NFTs.

• Social: finding myself spending more time on Orb.

• AI: unsubscribed from ChatGPT and currently using Venice AI and Heurist.

• Photography: taking verifiable photos on Click.

• Payments: Clave is amazing.

On Crypto’s Lack of Culture Market Fit: Crypto has many challenges to face (political enemies, lack of regulation, bad actors are just a few) but I hold the unpopular opinion that its biggest problem is that it still hasn’t found Culture Market Fit. By that, I mean that it still hasn’t convinced the younger generations that it’s not just cool but actually groundbreaking to use crypto. This video with Ohio State's commencement speaker being booed by mentioning Bitcoin is just an example of how much the average, non-tech person hates crypto. I don’t blame them. Crypto has a very bad reputation but more importantly has failed to eloquently communicate its values and ethos to the masses. We spent too much time trying to find utility but we forgot culture. This is a short screenshot essay I wrote about this.

A really insightful analysis on the current state of the AI industry. It’s quite funny that one of the points made is that distribution is king (to which I agree) and a few hours later the first rumors of an Apple<>Open AI partnership came out. Another key point is that Microsoft isn’t so well positioned anymore. This is also something that I agree with. A personal conclusion is that the Open AI<>Microsoft partnership is poorly designed and it will either backfire (almost did a few months ago) or will create conflicting market positioning. Will write a post dedicated on this soon.

Eli Dourado on how societies collapse. Exellent essay.

Hidden culture gem. Asterisk online magazine. It’s added in my list of favorite Internet corners with: Palladium, Noema, Works In Progress.

Anthropic just killed prompt engineering.

Climate activists attacking Tesla’s factory in Germany is the most post-surrealistic thing you will watch this week. It also has GenAI video vibes. It feels like Sora made it. Which makes the whole thing even weirder.

Money doesn’t solve homelessness. I have spent a lot of time reading and thinking about homelessness because it still strikes me as something which is relatively easy to solve for a country with the economic power of US. This report is excellent not because it comes up with amazing insights we didnt already know but because it outlines everything that has been already documented. Homelessness is a multi-variable problem that requires good data, strong coordination between public and private sector, money and most importantly a logical way of thinking. None of that exists in the US and especially in the West Coast.

Recommended podcast: Sam Altman on All-In Podcast. It’s very interesting to me how Open AI and Sam Altman aren’t so popular/loved anymore, especially in tech crowds. Of course, there are many people who don’t approve Open AI approach regarding AI regulation and OSS models and some others who got disappointed by the whole Open AI drama. But I believe it’s also the fact that Altman quickly converted into a political animal. Every word he says is extremely calculated (which I personally find OK considering his position) and unfortunately lacks of any major insight/substance. He is just extremely careful not saying anything out of script and ends up sounding like a bot. I guess someone would argue he isn’t a live player anymore.

Recommended Blog: Not Boring by Packy McCormick

Recommended Book: Deep Utopia: Life and Meaning in a Solved World Hardcover - Nick Bostrom

Quote: “Happinness writes in white ink in white pages”