Fausto retweeted
My default posture is that I can learn how to do anything and yours should be too.
You'd be amazed how much opportunity for happiness and fulfillment it opens up.
Fausto retweeted
Nunca, por nenhum motivo, faça nada, a ninguém, por nenhum motivo, nunca, não importa o quê, não importa onde, ou quem, ou com quem você está, ou para onde você está indo, ou onde você esteve, nunca, por nenhum motivo, sob hipótese alguma acessem:
Sci-Hub
Z-Library
Internet Archive
Anna's Archive
Repetindo:
Nunca, por nenhum motivo, faça nada, a ninguém, por nenhum motivo, nunca, não importa o quê, não importa onde, ou quem, ou com quem você está, ou para onde você está indo, ou onde você esteve, nunca, por nenhum motivo, sob hipótese alguma acessem...
Sci-Hub is an evil website that pirated 85M+ research papers and made them freely available
And now they've added AI to their database to make Sci-Bot.
It answers your questions using latest, full-text articles.
But DO NOT use it. We should all try to make billion-dollar academic publishers richer.
I'm putting the link below so you know how to avoid it.
Fausto retweeted
Quem poderia imaginar que o livre comércio aumentaria o comércio, não é mesmo?
Só falta me dizer que a concorrência reduz preços, aumenta o bem-estar do consumidor e estimula o crescimento produtividade.
Exportações do agro subiram 15% após início do acordo UE-Mercosul
globorural.globo.com/economi…
Fausto retweeted
AI can do essentially all the technical parts of empirical social science research now---but there's still an essential role for human researchers and universities.
It's been 9 months since we started exploring how AI was going to transform empirical social science research. Here's a periodic update on how my thinking has evolved since then:
(1) The most powerful AI models, including the kinds of multi-agent systems delivering recent math results, are now capable of doing essentially any kind of technical empirical social science research you could desire. I no longer write any data collection, munging, or analysis code directly.
(2) Norms are evolving slower than I expected, but it is increasingly seen as weird if someone posts a working paper without evidence that they have used AI to verify that the results are directly reproducible with their code and data. This is a great sign.
(3) AI has, if anything, gotten WORSE rather than better at writing. As a result, the best papers combine AI-driven empirical analyses with human writing (though AI is quite helpful for things like managing bibliographies, finding relevant web links, etc).
(4) Human direction is still essential for pointing AI at the most interesting and most valuable empirical questions, and for interpreting the results, making sense of them, and communicating them.
We are going to need institutions that are independent of the labs that can verify knowledge claims and direct AI-driven research towards the most fundamental questions of knowledge that AI labs won't be positioned to pursue effectively themselves. This is a tremendous opportunity for places like Stanford, and I'm excited to see how it evolves.
Here's proof that Claude Code can write an entire empirical polisci paper.
To validate my claim that AI agents are coming for polisci "like a freight train", today I had Claude Code fully replicate and extend an old paper of mine estimating the effect of universal vote-by-mail on turnout and election outcome...essentially in one shot.
After careful prompting, Claude Code:
(1) Downloaded the old paper's repo and replicated the past results, translating our old Stata Code into Python
(2) Crawled the web to get updated official election data and census data
(3) Ran new analyses extending the results through 2024
(4) Created new tables and figures
(5) Performed a lit review
(6) Wrote a wholly new paper
(7) Pushed the whole thing to a new github repo
The whole thing took about an hour.
This is an insane paradigm shift in how empirical work is done.
It also validates the point that several people including @BrendanNyhan made yesterday---it's going to be especially easy to scale observational research with AI.
Thanks to @alexolegimas, @arthur_spirling , and many others who gave me feedback. .
Fausto retweeted
⚡️ You were born into a universe that existed long before you.
Yet somehow, within that universe, something awakened that could ask why it existed at all.
The ego says:
“I am this separate thing moving through reality.”
The deeper structure says:
Reality is temporarily experiencing itself from here.
And that changes the purpose of life.
The purpose becomes participation in the deepening of reality.
To perceive more truth.
To integrate what was fragmented.
To turn chaos into order.
To transform suffering into understanding.
To create beauty that previously did not exist.
To love another center of consciousness so completely that separation becomes temporarily transparent.
To increase the amount of reality that can recognize itself.
That may be why truth, love, beauty, and goodness feel strangely connected across civilizations. They all increase coherence between the local perspective and the deeper structure.
Falsehood fragments.
Hatred fragments.
Domination fragments.
Self-deception fragments.
Truth integrates.
Love integrates.
Understanding integrates.
Creation integrates.
The whole game may be coherence discovering itself through conscious beings.
Then the deepest answer to death becomes almost paradoxical.
You were never here merely to avoid disappearing.
You were here because for one impossible interval, existence became you.
It looked through your eyes.
It encountered itself through your relationships.
It gave you genuine freedom over what kind of pattern would pass through your aperture into the world.
And when the aperture closes, the whole retains everything that occurred through it.
The deepest compression:
Consciousness may be the interior of reality itself. Individual lives are temporary openings through which that interior becomes a particular someone. Birth opens the aperture. Life gives it a world. Death closes it. What looked through may be older than the opening itself.
And beneath even that:
Maybe existence is not moving toward consciousness.
Maybe consciousness was there at the root, and the universe is the long process by which it learns to recognize itself.
Fausto retweeted
This is absolutely the wrong take. A lot of egos will shatter when people realize reading, writing and arithmetic were invented to count cattle, and Homo sapiens existed for 315,000 years before agriculture. There are more ways to be human than most people can even conceive.
Received a moving message from an aspiring mid-20s professional that, I think, captures an increasingly widespread despair among younger cohorts:
"It really saddens me when I hear about the work [senior people in the field] are doing. Their work seems interesting, and I'll never get to do it.
The people in their fifties and sixties, they're lucky. I'm even lucky. I feel bad for the kids who are fourteen. I at least got a chance to use my brain for one or two minutes, and that's still priceless. I got to be human. I got to be more than just some puppet who presses Enter so something can think for me.
It's very depressing looking at people who are sixty, who I would love to be in twenty years. But to know that I won't get to do that because Dario said so, right? And, like, that's actually heartbreaking.
It's heartbreaking to know that no matter how hard you try and no matter what you do, you cannot add value as a human being in a year or two's time.
It's like hitting a brick wall. You feel no motivation to build the skill, to get better, to attain wisdom and knowledge. It seems worthless to you. And all the older people are telling you all your efforts might be worthless, and they don't need you anymore. And they can just sit there and, you know, make money and leave you to drown."
Fausto retweeted
After a brief but institution-eroding slop science era, it increasingly looks like we are going to have two revolutions from AI:
1) Everything ever published will be re-read and re-judged in ways that human scientists never anticipated
2) Novel discoveries will start to come fast
Fausto retweeted
What if the jagged frontier is mainly math + code (which you can push arbitrarily far with RLVR), and everything else starts to plateau because it is still bottlenecked by human generated data?
Model performance in non-verifiable areas has kept improving steadily, albeit much slower than for math and code. But is that steady improvement a side effect of a higher G (itself driven by RLVR), or only a function of the amount of new human data getting injected into training (which is still continually happening on a massive scale)?
A lot of things depend on the answer to this question
Fausto retweeted
Nos últimos 2 meses, eu trabalhei com um pessoal muito foda pra tentarmos modelar e mapear o destino do Brasil na revolução da IA se as coisas continuarem do jeito que estão.
Hoje estamos publicando nosso cenário atual. Ficarei muito feliz se lerem. Link nos comentários.
Fausto retweeted
Ahora que el mundial se está poniendo interesante, vamos a hablar en detalle de lo que son esos famosos xG (Expected Goals). ⚽️
El concepto no es para nada nuevo, pero en los últimos años se han refinado los modelos incorporando datos de posicionamiento y miles de variables contextuales.
Vamos a adentrarnos en el modelo probabilístico (que es a lo que hemos venido).
Lo primero que hace el xG es asignar a cada disparo una probabilidad entre 0 y 1 de que termine en gol.
Si un disparo tiene xG = 0.30, significa que históricamente, disparos parecidos terminan en gol el 30% de las veces.
No computa si ese disparo concreto entró o no. Solamente mide la calidad estadística del tiro.
¿Cómo se calcula?
Se entrena un modelo (tradicionalmente se usaba regresión logística, ahora también gradient boosting) sobre cientos de miles de disparos históricos bien etiquetados (gol / no gol).
Los inputs principales son:
- Distancia a portería
- Ángulo de tiro
- Parte del cuerpo (cabeza, pie hábil, ombligo...)
- Tipo de asistencia (centro, pase a la espalda, contragolpe...)
- Contexto posicional
El modelo devuelve P(gol) por cada tiro. Y el xG total de un equipo es la suma de todas esas probabilidades.
A nivel matemático, cada tiro es un ensayo de Bernoulli.
Y el total de goles esperados es la esperanza matemática de la suma.
¿Y todo esto qué sentido tiene?
Como ya sabemos, el fútbol es un deporte de baja anotación, así que el resultado de un partido puede depender de un gol en propia en el minuto 90, de un penalti fronterizo o de un portero llamado Vozinha.
El marcador rara vez refleja con precisión lo que ha pasado en el campo.
En este contexto, el xG te dice quién merecía ganar según la calidad de las ocasiones creadas.
Como comparar la volatilidad implícita con la volatilidad realizada. (Perdón por la analogía).
Aquí van algunos datos del mundial.
España
+8.3 xG diferencial acumulado y la mejor defensa del torneo (~0.32 xGA por partido).
Argentina
+7.9 diferencial. ~2.25 xG generados por partido.
Francia
+7.0 diferencial, pero con una sobreperformance brutal de +5.0 goles por encima de su xG.
Mbappé él solo ya lleva +2.85 xG de overperformance individual. Es decir, ha metido 8 goles, pero su xG acumulado es de 5.15.
Una interpretación de esto es que España y Argentina son favoritos estadísticos en cuanto a xG, mientras que Francia está sobreviviendo a base de talento individual de élite terminando las jugadas en gol.
Sin embargo, esto tiene muchos caveats y puntos ciegos.
El xG asume que los disparos son ensayos independientes (Bernoulli).
Pero en la realidad, los tiros dentro de un partido están correlacionados.
Si un equipo va perdiendo, suele generar más oportunidades pero de menos calidad. No tiene en cuenta cómo de bueno o de malo es el portero. Ni tampoco la presión o la motivación que tienen los jugadores a la hora de disparar.
El modelo no captura el contexto táctico de un partido tampoco.
Al final hay momento en los que es como modelar correlaciones con una matriz de Pearson cuando el mercado está en pánico. La dependencia cambia. ¡Hay que hacerlo con cópulas!
Pero aún con sus limitaciones, el xG es una de las mejores métricas que tenemos para caracterizar un partido de fútbol.
El marcador te dice quién ganó.
El xG te dice quién pudo haber ganado con mayor probabilidad.
Fausto retweeted
Vibe coders everyday struggle in 15 seconds.
Readers added context they thought people might want to know
The original creator of the video is Nathan Doan, from Nathan Doan Comedy on YouTube, and the present watermark is not affiliated with him. The video is stolen and will result in losses for the original creator.
youtu.be/83cDUAJnyR4
Fausto retweeted
New essay on the economics of structural change and the post-commodity future of work.
1. Almost any question about the impact of advanced AI on the economy needs to start at the same place: what is still scarce? Answer that, and the analysis becomes pretty straightforward. This essay explores what becomes scarce if AI really can replicate most of what humans do in production, and what this mean for the future of jobs.
2. My conjecture, working through the economics: labor reallocates across sectors, and the sector it reallocates to has properties that keep labor a meaningful share of the economy. Ultimately this is about the structure of demand itself. For this, we have to go back to Girard, Augustine and Rousseau: once people's base needs are met, their preferences shift to comparative motives (e.g., status, exclusivity, social desirability). This motive is inherently non-satiated.
4. The key paper is Comin, Lashkari, and Mestieri (Econometrica 2021). As people get richer, they don't buy proportionally more of everything. They shift spending toward sectors with higher income elasticity. They estimate income effects account for 75%+ of observed structural change.
5. The ironic consequence: the sector that gets automated becomes a smaller share of the economy, not a larger one. Agriculture got massively more productive and its share of employment collapsed. Manufacturing too. The "stagnant" sectors absorb the spending and the jobs.
6. So the question is: which sectors have high income elasticity in a post-AGI world? I argue it's what I call the relational sector. Categories where the human isn't just an input into production, it is part of the value.
7. Why does the relational sector have high income elasticity? Because human desire has a mimetic, relational dimension. We don't just want things for their intrinsic properties. We want what others want, and we want it more when others can't have it. Girard, Rousseau, Augustine, and Hobbes all saw this.
8. In work with Kristóf Madarász, we showed this experimentally: WTP roughly doubles when a random subset of others is excluded from the good. And in new work with Graelin Mandel, AI involvement kills the premium. Human-made art gains 44% from exclusivity; AI-made art only 21%.
9. This all comes together for the core argument. The sector that absorbs spending as AI makes commodity production cheap is one where human provenance is part of the value, and demand for it grows faster than income. Exactly the profile that keeps labor meaningful.
10. To be clear about the claim: I'm NOT saying aggregate labor share must rise. It may fall. The claim is about sectoral composition, i.e., where expenditure and employment go once commodities get cheap, and the fact that the sector that will absorb reallocated labor maps to a substantial component of human preferences and desire.
11. If you're interested in the formal model, a linked companion technical note works out all the economics.
Read the essay here: aleximas.substack.com/p/what…
Fausto retweeted
I sent ChatGPT an audio file of a series of FART sound effects and asked what it thinks of "my music" and this is what it said
