Nico D (@srdipi) header
Nico D (@srdipi) profile picture

psiconauta del inconsciente.

Joined April 2012
Photo by Nico D (@srdipi)Photo by Nico D (@srdipi)Photo by Nico D (@srdipi)Photo by Nico D (@srdipi)Photo by Nico D (@srdipi)Photo by Nico D (@srdipi)Photo by Nico D (@srdipi)Photo by Nico D (@srdipi)
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel. - Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all. - Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract. - Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code. - And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while. The funnel is driven by a combination of factors: 1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel. 2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel. 3. Access. Free-tier, paid-tier, internal. So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code. But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along. So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions. TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.
2,266
5,940
2,297
44,375
7,017,946
Nico D retweeted
En este océano de información dudosa, hay que escuchar e informarse por lxs que saben y son serixs 👇👇👇👇👇
Para saber si un investigador principal gana más de 4 millones podemos ver la web de @CONICETDialoga conicet.gov.ar/transparencia… El sueldo bruto promedio de un principal en febrero 2026 fue $3.184.731 ($ 2.722.000 en mano). Desde ese momento subió un 12% (3 millones en mano).
5
33
108
5,032
Nico D retweeted
"Guarda que el que no salta es un inglés pero tenemos unos criollos que nos están haciendo pelota" Los HUEVOS de Ricardo Mollo, artistas de verdad bancando al pueblo, hermano
71
1,691
30
10,605
84,027
Hoy, Estela cumple 95 jóvenes años. Ayer por la tarde estuvo en San José 1111. Vino a visitarme. La serenidad y calidez que trasmite siempre fue algo que me llamó la atención. Supo transformar su pérdida y su dolor en obra colectiva y reparadora. Sin lugar a dudas, Estela es una de esas personas imprescindibles en la historia de nuestro país. Feliz cumpleaños… Estela de Carlotto.
2,641
4,466
304
26,139
883,645
“De ninguna manera le voy a desear un feliz cumpleaños a un sorete perverso y cruel. Para no cumplir la Ley de Discapacidad tenés que ser un verdadero hijo de puta” ✍🏻 @Lipovetzky en #DesayunoIntermitente
47
127
3
1,174
29,269
¡Habla bien LCDTH! 😆😂🤣
100
1,400
69
5,441
90,434
“Nada beneficia a Argentina, se estan muriendo, no tienen plata, no tienen nada, me cae bien su Presidente pero no lo hagas sonar como que lo están haciendo bien, se están muriendo”.
697
3,434
553
18,107
1,023,017
No bueno digo, pero digamos si bap, digo vamo' a deci' no, total no se puede no se que ueep digooo..
55
545
30
3,726
63,221
🥹 déjenme sola
2
24
5
98
3,461
Lo que jugó la bomba tucumana ayer en la mesa de Mirtha fue impresionante. Con Cristina Pérez sentada enfrente vale doble.
276
1,066
39
7,797
107,150
Pérez trata de usar el caso del hijo de Valentina Bassi, que es autista, en favor de su argumento falaz. Y Bassi le da una clase de periodismo, sin ser periodista. Brinda datos, sale de su caso personal, analiza contextos y consecuencias. Touche.
Por favor no se pierdan la cara de Cristina Perez. Para alquilar balcones. Es la cara de una mujer totalmente derrotada, en pocas palabras: Valentina Bassi le pego un guantazo de realidad en el medio de la jeta.
42
771
8
4,545
62,232
Nino Bravo cantó "Libre" para hablar sobre la represión a los jóvenes durante la dictadura de Franco Charly compuso "Demoliendo hoteles" para criticar la represión de la dictadura empresario militar y los valores que imponían a tiros y torturas. Milei no entiende ni lo que intenta cantar. No tienen ídolos, ni canciones, tienen que robar las nuestras. Hay que cantarlas más fuerte.
519
3,688
98
18,936
531,537
Con Milei la única industria que crece es la de la mentira. La foto mía es falsa, no fui de compras, no me corté el pelo, no llevé un traje negro a EEUU. Pero la imagen de Milei entregando la política económica a Trump es tristemente verdadera. Dejen de estafar, de mentir y de fabricar fakes que no causan gracia. Póngase de una vez a gobernar en favor de Argentina.
2,639
6,414
481
29,186
3,253,294
¿SE PUEDE SER MARADONIANA Y FEMINISTA?
84
125
46
1,850
99,193
Momento en que los manifestantes de la Marcha Ni Una Menos al ser reprimidos por las fuerzas del régimen se defienden y los haces salir corriendo
349
1,522
130
9,090
185,596
Imaginate ser gorila y esperar años para verla presa, y cuando por fin lo consiguen ELLA ESTÁ ASÍ
903
2,257
142
28,056
552,816
Pueblo 2 - Milei 0 (Karina 3%)
1,019
3,764
119
25,519
449,164
ME PONGO UNA CAMPERA ME PONGO UNA CAMPERA Y ARRIBA OTRA CAMPERA NECESITO UNA CUARTA CAMPERA.
25
636
35
6,278
155,055
"Pelotudo" Por esto que dijo Pedro Rosemblat sobre Iñaki Gutiérrez.
315
1,422
229
17,886
912,090
El dato que le pasaron a Milei, según Majul, es incorrecto. Si se “tranquilizó” con eso, como dice, se tendrá que intranquilizar de nuevo. La Andis nunca dependió de la Secretaria General de la Presidencia en esta gestión. Cuando se creó, en 2017, estaba en esa órbita. En 2019 pasó a Salud y permaneció ahí hasta 2023. Cuando llegó Milei, la sacaron de Salud y la pasaron a Jefatura de Gabinete (Decreto 45/23). En julio de 2024, cuando ya se había ido Posse, volvió a Salud (Decreto 585/24). Tardaron literalmente cinco días en inventar una explicación para tratar de despegar a Karina y la explicación es sencillamente mentira. Están perdidos con este quilombo.
Último momento. Le acaban de pasar a @JMilei un dato "tranquilizador". La @andiscapacidad hasta diciembre de 2023, dependía de la Secretaría General de la Presidencia. Pero @KarinaMileiOk dispuso que pasara a la órbita del @MinSalud_Ar. "Si realmente hubiese querido afanar ¿por qué no quedarse con la caja, en vez de entregarla?" se preguntaron cerca del presidente.
562
1,875
116
10,916
783,532