Simón Muñoz

Principal Product Manager. Tweets sobre producto, tecnología y startups.

[Brandon Carl @brandonjcarl]

Documenting the headwinds I now see for AI.

It won't seem like it, but I love AI and am long-term positive. But when "math doesn't math" I take note.

1. The core thesis for foundation model lab investment has been high upfront investment made worthwhile by significant long-term profits.

2. These are capital intensive businesses and the compute commitments are very high relative to revenue and require strong growth over long time periods. The "leverage" (commitments versus revenue) is extremely high.

3. The fundamentals are not as positive as they previously were:
• Input costs are higher (commodities, chips, power)
• Interest rates are higher
• Competition is more intense
• Scaling Laws are now problematic: exponential costs/power cannot continue

4. Forecasting compute spend is challenging and high risk due to (a) revenue uncertainty and (b) algorithm uncertainty

5. Revenue growth appears to be slowing. The technology is valuable, but ROI is proving to be more expensive and take longer than anticipated.

6. The future is likely "different models for different use cases" with the lower end of the market being highly competitive.

7. Core use cases such as agentic software engineering are likely to need approaches beyond next-token prediction. They are Σ₂ᴾ complexity problems requiring multi-objective optimization and likely a combination of Transformers and other methods.

8. Current forecasts in memory makers are built largely on quadratic attention. That will not persist: we are already seeing work from DeepSeek, Minimax and Nvidia that can cut RAM needs by 80% or more.

9. This means semiconductor valuations are substantially overinflated and will go through the traditional glut versus shortage cycle.

10. For foundation model providers: lower costs with competitive differentiation is good. However, lower costs with a lack of differentiation would mean lower revenues. This makes it harder to (a) service commitments and (b) pay back investors.

11. Leverage is substantially higher than in previous cycles, evidenced by leveraged ETFs, call option activity and margin loans. Korea is particularly susceptible.

12. 0DTE options create a profile that has stronger parallels to portfolio insurance and 1987 than any other point I can remember.

13. The combination of exponential increases in call activity coupled with the ties of semiconductors to structured products means there is a non-trivial systemic risk to the financial system.

14. Implied earnings growth rates are inconsistent with other periods in history.

15. Macroeconomically we cannot and should not fund exponential cost increases. History has shown us repeatedly that there are better ways (see Quick Sort and Simplex).

16. Significant supply is hitting the market via IPOs.

––

Taken together: costs and competition are increasing while revenue growth is likely slowing. Valuations are fragile and prone to technology disruptions that are already here. Systemic financial market risk is extremely high.

[𝗠𝘂𝗵𝗲𝗲 ♛ @muheediva01]

VERY unpopular opinion: It’s okay to NOT want to be a manager. You can have a fulfilling career and never manage a single person.

[Uncle Bob Martin @unclebobmartin]

There is no way to get the human out of the loop. The Agents simply do not understand obvious user expectations.

[Simón Muñoz @simonvlc]

@davidalvarezdlt

[Compound248 💰 @compound248]

If I am @sundarpichai, I will bury all competing foundational models in negative gross margin tokens for eternity.

I will NEVER take my foot off the pricing gas: to merely compete, I will force you to dilute your equity into oblivion.

Sam will drown in my tsunami of COGS. $GOOG

[Klaas @forgebitz]

everyone is now suspiciously: "ai will not replace jobs"

complete 180 within 2 weeks

[grk @monyequest]

todos no podéis tener el síndrome del impostor algunos tenéis que ser incompetentes de verdad

[𐌁𐌉Ᏽ 𐌕𐌉𐌌𐌉 @OrevaZSN]

2016: I use technology every day, and I love it.

2026: I want a job as a park ranger who only needs a flip phone.

[Simón Muñoz @simonvlc]

En Every, pese a haber automatizado al máximo su trabajo con agentes, no han despedido a nadie y tienen más trabajo humano que nunca.

Las empresas que de verdad aprovechan la IA no solo no despiden, sino que contratan más.

https://t.co/ABXBRNIlwS

[jezz @JezziiB]

Tired: Humans are killing the planet

Wired: literally like 90 specific billionaires and CEOs are killing the planet and we literally know their names

[Malika 🧬 @malikules]

atheism leaving my body with every new Pope tweet

[Tim Duy @TimDuy]

I just can’t see how firms decided to reward employees for maximizing use of inputs rather than maximizing output.

[𐌁𐌉Ᏽ 𐌕𐌉𐌌𐌉 @OrevaZSN]

AI seems smart when you ask it about things you know nothing about, but once you start asking it about topics you are actually an expert in, you realize it is just commoditized, polished, and amplified mediocrity.

[Simons @Simon_Ingari]

The final stage of your career is being silent in meetings.

[Andrew Gazdecki @agazdecki]

Founders finding out building a startup is 90% sales and marketing:

[Simón Muñoz @simonvlc]

“CEOs are uniquely prone to AI psychosis because they’re sufficiently distant from the last mile of work that still has to happen to generate most value with AI”

[TechCrunch @TechCrunch]

Tech CEOs are apparently suffering from AI psychosis https://t.co/AuRuruC5rO

[Martinez @MrtnzAlvrz]

What +100% annual returns do to a country

[Boze Herrington, Library Owl 😴🧙‍♀️ @SketchesbyBoze]

The internet used to be full of websites; there were millions of them and you could browse for hours and come away smarter rather than dumber. Now there are four sites and they've made half the population illiterate. We've destroyed a wonderful thing, and it has destroyed us.

[Stefano ₿Ξrnardi 🌲🏴 @stefanobernardi]

The Fiat Multipla version is better ☠️