PassiveInvestor420 🙈⏳

I was recently reminded that fruit of the loom was one of the least admired companies of 2000-01 by Fortune magazine, along with Trump Hotels and Casinos. Among the most admired? Enron.

@karpathy On-demand documentaries please!

This works really well btw, at the end of your query ask your LLM to "structure your response as HTML", then view the generated file in your browser. I've also had some success asking the LLM to present its output as slideshows, etc. More generally, imo audio is the human-preferred input to AIs but vision (images/animations/video) is the preferred output from them. Around a ~third of our brains are a massively parallel processor dedicated to vision, it is the 10-lane superhighway of information into brain. As AI improves, I think we'll see a progression that takes advantage: 1) raw text (hard/effortful to read) 2) markdown (bold, italic, headings, tables, a bit easier on the eyes) <-- current default 3) HTML (still procedural with underlying code, but a lot more flexibility on the graphics, layout, even interactivity) <-- early but forming new good default ...4,5,6,... n) interactive neural videos/simulations Imo the extrapolation (though the technology doesn't exist just yet) ends in some kind of interactive videos generated directly by a diffusion neural net. Many open questions as to how exact/procedural "Software 1.0" artifacts (e.g. interactive simulations) may be woven together with neural artifacts (diffusion grids), but generally something in the direction of the recently viral x.com/zan2434/status… There are also improvements necessary and pending at the input. Audio nor text nor video alone are not enough, e.g. I feel a need to point/gesture to things on the screen, similar to all the things you would do with a person physically next to you and your computer screen. TLDR The input/output mind meld between humans and AIs is ongoing and there is a lot of work to do and significant progress to be made, way before jumping all the way into neuralink-esque BCIs and all that. For what's worth exploring at the current stage, hot tip try ask for HTML.

@buccocapital I personally felt exactly the opposite during the social media era. Like I was trading my data for the privilege of vanity posts and targeted ads. This time I get something truly useful in return and it’s amazing to see it advance every year.

The vibes are turning against Dario. I expect this to accelerate until he changes his messaging Why? Because AI broke the social contract of the internet. For the last 20 years, people shared their data with tech companies as part of a *symbiotic* relationship. The services got better. Traffic went to your site. Network effects made transactions liquid and simple. AI broke that contract. AI sucked in the collective IP of civilization and paid out nothing for the inputs. And now Dario goes on TV every day to tell us he is going to break the economy and our children will be poor. But he did that. It’s economic gaslighting. Dario is the embodiment of the “We’re all trying to find the guy who did this” meme and people are not going to sit there and take it AI needs a better, more inclusive message. I expect he will realize this eventually. The question is whether he will realize it before it’s too late.

@DimitryNakhla What zuck’s saying is a little different and too rosy: that companies’ resources can be moved from marketing to product improvement. It’s more liekely to be captured as revenue, and that’s assuming the marketing costs actually go down (Meta controls pricing)

Mark Zuckerberg subtly explains how $META advertising moat is widening due to AI. The old model was simple. Demographic targeting based on profile data that users voluntarily input. Today, it’s much different. Meta’s AI now generates up to 4,000 versions of your creative, tests them in real time, and identifies who actually wants your product — often more accurately than the advertiser themselves. Zuckerberg’s message to advertisers? Stop constraining the AI. Give it a hint, then let it work. This matters a lot for the business model. When advertisers get better outcomes — higher conversion rates, lower cost per acquisition, stronger ROI on every dollar spent — the value of that ad spend becomes easier to justify. And when the value is easier to justify, $META has more pricing power on top of growing demand. Not because they’re forcing it, but because they’ve earned it. Better outcomes → higher advertiser ROI → increased ad spend per advertiser → greater pricing power for $META → expanding margins. And layered on top of all of this is a distribution advantage that is nearly impossible to replicate. The scale at which $META can run these experiments — across billions of users is simply not available anywhere else. ___ 🎙️ Stripe | A conversation with Mark Zuckerberg (05/08/2025)

@buccocapital While I agree and this tracks, you present context as a finite resource. Call me an optimist, but why shouldn’t we expect context discovery to be a value-add new job description?

Claude Cowork is blowing my mind. But it’s only as good as the context it can gather The race is on to capture every piece of context in the organization Companies will use carrots and sticks to get that context out of their worker’s brains. Work is about to get very weird

@awealthofcs I wonder why BTC is going sideways (vs Nov ‘24 valuation) as opposed to falling further if it’s perceived role isn’t as expected

This was the perfect set-up for crypto: -Faith in the system waning -Pro-crypto admin -$ is falling -Fed independence in question And yet...investors still prefer physical gold to digital gold Score one for the goldbugs

@awealthofcs You guys killed it with the Jurien segment

Animal Spirits: All-Time Highs Should Feel Better Than This Ex-Mag 7 stocks are breaking out No one wants a bubble Stock market returns in 2026 Howard Marks on hedging AI Data centers in space? The 2026 housing outlook & more awealthofcommonsense.com/2025/12/animal…

@DataTrekMB Please make a video about this!

Since 1930, the fifth year of a given decade has averaged exceptionally strong S&P 500 returns. This year has not followed that pattern, and the 2020s is proving to be an unusual decade in other ways as well.

@buccocapital Revenue models same as it ever was.

A new paper from researchers finds that GenAI-created ads see 19% higher click-through rates than human-created ads $META $GOOGL the AI winners.

@buccocapital What companies can withstand both an implosion in demand and an Industrial Revolution-style increase in productivity due to economy-wide reduction in payroll overhead? It’s an extremely narrow distribution of winners.

The market is pricing in a few companies sucking up almost ALL of the value of AI: - Chips - Semicap + Energy - Platform choke points (AAPL + GOOGL) - Cloud infrastructure This is obviously and hilarious not true, but it’s all that’s visible so people are gunning these stocks 5% every day But! There’s a glimmer of people actually starting to consider where else value might accrue. Healthcare/Bio starting to get bid, for example. Some robotics stuff. Ads definitely a winner, though already bid But worth thinking through where the actual value will accrue, because the pricing in the market, IMHO, cannot be the actual end state Someone has to make money for this work or it all blows up. Nobody will spend on AI and it will just…blow up. The money can’t all go to the inputs. So who will make money from AI? That’s the game for 2026.

@buccocapital LLMs have squeezed all the juice from the era of scraped content. Hyperscaling has created a huge appetite for fresh training datasets, and in-the-field sensors are an untapped area for satisfying that need.

@buccocapital I think it’s worth differentiating Amazon the marketplace and Amazon the cloud service, if only for the purpose of clarity here

Added more Amazon today and started to build my position in Meta Boring, I know But I continue to have conviction that AI adoption is: 1. Very real and genuinely valuable 2. Mostly a sustaining innovation 3. A game of kings 4. Going to be won through distribution and scale Nobody is better positioned than the greatest businesses on earth This past week OpenAI tipped their hands on concerns that are idiosyncratic to OpenAI themselves, and in fact made me incrementally more bullish on the incumbents to win

@ConradEduard @buccocapital Despite volatility will continue to move up and to the right

@buccocapital "Mostly a sustaining innovation" Can you elaborate a bit in this?

@christianbrits @CNLiberalism Wouldn’t a better slogan be “zoning reform?” This has a draconian tone similar to “defund police.”

.@CNLiberalism you need to warn people your stickers will get people enhanced scrutiny at airports

Just got into an argument with the TSA gate agent about my ‘abolish zoning’ laptop sticker. Without zoning, he says people could just build anything 🤷‍♂️

@themarketuni Coming from a numbskull such as myself, I wouldn’t be borrowing the quarter before multiple rate cuts unless I absolutely had to.

US businesses are pulling back on borrowing. Q2 2025 saw the weakest demand for commercial & industrial loans since late 2023, for both large and small firms. Such low appetite has only appeared in 2020 and after the Fed’s 2023 rate hikes — a clear sign loan activity is fading.

This depends on the degree to which recursive self improvement played a prominent role in training this new model, and how this momentum can be followed across all AI development models.

In “Superintelligence” (2014), Nick Bostrom suggests, “attempts to build [AGI] might fail… until the last missing piece is put in place.” Ultimately, GPT-5 may emerge as an important moment in iterative, recursive self-training for future AI leaps forward.

GPT-5’s release has been met with mixed reviews. Many users are disappointed in its lack of meaningful improvement over previous models. However, it’s possible the salient issue is the prevalence of “recursive self improvement” in its development.

mark zuckerberg is buying researchers for as much as a billion dollars a pop to build a sand god, and the world is talking about tariffs.

@DataTrekMB Very fair point. Non-correlation and diversification are not mutually exclusive. Shout out to the DataTrek newsletter. I look forward to it each day it comes out.

@oc_mooch Correlations can temporarily spike but few things in a crisis are a hedge when things go south. Our chart covers the last year and a half. We use 30 days because it catches spikes like the tariff scare.

1/2 #Bitcoin only shows a 30-day price return correlation of 0.39 to the S&P 500 & Nasdaq 100. That’s an r-squared of just 15%, meaning that the price action in stocks – Tech or otherwise – only “explains” a small part of $BTC's changes in value...

@DataTrekMB It will have to continue this trend well past 30 days to truly offer diversification. We’ve already seen it fall in concert with stocks during the tariff scare and rate hike shock

2/2 ...It also trades independently of gold, with 30-day correlations to #bitcoin of only 0.11 (r-squared of 1%). $BTC continues to chart its own path, and therefore offers diversification benefits to a portfolio of financial assets.