Sometimes macroeconometrics by day, always maths by night (or viceversa) and algorithmics somewhere in between. Art lover. Tweets are my personal opinions.
Joined February 2017
- Tweets 26,369
- Following 1,265
- Followers 1,184
- Likes 20,208
2,837 Photos and videos
Jun 5
Otro grande que se nos va. Si habre estudiado con sus libros, esp. "Parallel and Distributed Computation" con John Tsitsiklis y "Nonlinear Programming". @wsosaescudero
Jun 5
I've just learned that Prof. Dimitri Bertsekas @DBertsekas has passed away.
His contributions to optimization, dynamic programming, control, and reinforcement learning shaped generations of researchers and engineers. I am grateful for the knowledge and insight I gained from his books and work over the years.
My sincere condolences to his family, friends, colleagues, and students. His impact on our field will be felt for generations.
1
44
May 30
Testing, yet again, frontier models to execute meaningful analysis of rather complex codebases. Something that was bugging me. Thought "maybe I'm missing something here" and had a go. But, but...
It took at stance. I said "can't be, it does not make sense." Doubled down: "you are wrong, I am right." Put pen to paper: "this can't be, but let's try, maybe I'm missing something." Tested empirically: the proposed solution was wrong, as I predicted. The answer I got was "you are wrong, if it worked before it was by accident, and my fix stands" type of response.
Then I blew the gasket, on purpose. The answer coming back is telling.
But why? Frontier AI tend to get lost (fail to see the wider coding context) beyond, say, 200-300 lines of code at either side of whatever you're looking at. But also, and more importantly, you need swearing at them with full strength (see screen cap) to avoid post-rationalization and bullshitting. That's even more dangerous.
If you know what you're doing and why it gives you a foot to stand, but if you do not (as most vibe-coders, definitionally) you swallow whatever bullshit is thrown at you.
This is why vibe-coding and "the death of software engineers" is just gaslighting. That's why vibe coding stuff will invariably break things. The more vibe-coded, the more complex it will be to backtrack. That's also why it'll likely become a compounding problem.
Granted, frontier AI can help you with certain things, but you'll end into a deep and dark rabbit hole if you let it loose with even mildly complex stuff.
Receipts provided. Things to note:
- I did bully Claude to shake it out of a post-rationalization loop. That's always nearly necessary, but it does not guarantee you'll land in a better place.
- Bullying has side-effects: Frontier AI models tend to capitulate easily to avoid alienating the user (you) even when the user is wrong (without knowing).
- Consequently, if you really don't know what you're doing, you end in between a rock and a hard place. You need to swear at them to break their COT out of self-confirmatory loops, but if you don't know your stuff shouting will lead to appeasing (the seemingly default response when swearing at them; this is not uniformly the case though) which may end up potentially confirming your own biases.
- That's why, once again, vibe-coding stuff is potentially lethal in production environments.
@GaryMarcus @rohanpaul_ai @Scobleizer @rauchg @wsosaescudero
2
39
May 23
There is a deep why: LLMs and all traditional DL models cannot deal natively with information from unstructured environments. They need to form first a map, a landscape, encoding it through the model weights. Lacking any radical innovation in the underlying tech stack, this is the only hope (and a blunt mechanism at it) for traditional models to become relevant, e.g. for household robotic assistants. @Scobleizer @GaryMarcus
1
109
May 8
Lady Lemon cosplay..? 🤣
Chau, cerra el estadio. En el horno con papas. @DanielMontoya_ @ladymarketok
x.com/i/status/2052485744447…
May 7
NUNCA ME OLVIDO DE ESTA JOYITA JAJAJAJAJAJAJAJAJAJAJAJA
Darth Vader:
"vengo de una galaxia lejana...
tirando a Merlo"
JAJAJAJAJAJAJAJAJAJAJAJA
2
346
Ed Salazar retweeted
May 2
Spirit Airlines died tonight at the hands of the socialist crusader, Elizabeth Warren
She must be so proud to add another casket to her achievements.
Tonight at 3am, Spirit turns off the lights. 14,000 jobs gone. 30 smaller airports lose service.
JetBlue offered $3.8 BILLION in cash to buy Spirit in 2022. Shareholders, flight attendants union, literally everyone voted yes.
The combined company would have held 9% of the US market against a Big 4 that already owned 80%.
For anyone who understands numbers: 9% isn’t a monopoly against 80%.
Warren said no.
She wrote letters. She pressured Buttigieg. Biden’s DOJ sued. A federal judge killed the deal in January 2024.
Her argument: the merger would cost consumers $1 billion a year.
Now look at her collateral damage she dusts under the rug.
510 pilots gone in the months after. 1,800 flight attendants furloughed in December.
14,000 jobs in 2023. 7,500 last week. Zero tonight.
And that’s just the people in Spirit uniforms.
Catering goes. Fuel guys go. Baggage crews, gate agents, airport coffee shops, hotels and rental cars in 70 cities Spirit flew to. Every airline job carries 3 more on its back.
40,000 people out of work because of one woman’s moronic crusade against the market.
And the math ain’t mathing.
Spirit abandoned 90 routes during the death spiral. Fares on those routes are up 14% on average. Oakland to Newark: $135 to $288. Fort Myers to San Juan: $92 to $219. Kansas City to Newark up 66%.
That’s reality. Not some BS number from a “study.”
So @SenWarren tell me how this saves the consumer money?
Cheap carriers in a market drop fares 21% across the board. Southwest did this in the 90s and saved Americans $68 BILLION over 20 years.
Warren killed it. That’s what moronic politicians led by socialism do.
Then with her own blind arrogance, she tweeted Spirit’s collapse is “a Biden win for flyers.”
A win.
14,000 people are reading termination letters tonight.
And she’s taking credit.
This is socialism in 2026.
A senator who’s never made payroll thinks she knows how to run a market better than the people who own and work in the company.
She saved you a billion on imaginary paper.
She cost you ten times that in real life.
She didn’t protect consumers from anything.
14,000 will go from working to welfare.
She will make sure to blame billionaires, hardworking tax payers, AI, capitalism and whatever monster they will make up tomorrow hiding under your bed.
Higher taxes. Fewer jobs. More expensive everything.
She called it a win. I hope you enjoy winning.
5,884
33,344
124,170
6,606,549
Apr 26
I've head stupid things, but this is pretty much top 5, even podium contender.
Anthropic CEO (Dario Amodei):
"Coding is going away first, then all of software engineering."
What do you think about this?
1
44
Apr 26
Al final, a los pibes se la pusieron de dorapa. Unos tiernos. Primero cantaban y bailaban al son de "Abdel Nasser" en la batuta del ex-gordo @CarlosMaslaton (parece ya como decadas atras) y ahora la tortera les termino de medir el aceite. Y que queres, si ensalzas de referente a Pepita Biscochuelos? Ademas de comerse a Adorni, A-DOR-NI, el 4 de copas (o florero) infaltable, y siguen las firmas. Mamadera. Masla, al fin y al cabo y cual piedra, fue profetico en el yunque que les iban a colgar al cuello a los pibes. Ahora a llorar al campito. @ladymarketok @VascoAmonda
Apr 26
El loco tenía lo que todos los políticos se desviven por tener.
Y lo prendió fuego para priorizar a unos cirujas que son el descarte del descarte del descarte de lo más rancio de lo que se votó destruir.
Será un experimento espectacular.
Económicamente en el camino correcto.
Políticamente envenenado.
1
107
Apr 23
Interesting paper. Is anyone surprised..? @GaryMarcus
x.com/i/status/2046997740559…
Apr 22
Let me say this clearly: LLMs cannot feel emotions.
Emotions are evolutionary mechanisms.
They push us to avoid danger or approach what is beneficial.
We experience emotions because we are alive, and we want to stay alive.
LLMs are not alive.
Yes, emotional language may be encoded somewhere in the LLM.
Yes, it may even be associated with some LLM output.
But that is just a superficial property.
There is nothing deeper behind it.
For a very simple reason: LLMs do not have an intrinsic and inescapable drive to stay alive.
This is what we call “motivation fault line” in our paper describing seven fault lines between human and artificial intelligence.
*
Paper in the first reply
1
34
Apr 23
Iranian trolling supreme @INArteCarloDoss @ParrotCapital
Not sure anyone read all of it yet. Does it have anything to do with us?
1
2
157
Apr 18
Just seen this and will read the paper later, but suffices to say the degree of cultism and post-rationalization in the replies is off the charts. If anything, a very human trait 😁 @GaryMarcus
A Google DeepMind researcher published a paper arguing that AI can never be conscious. Not a matter of time or scale. A matter of category.
The argument is that computation is a description of a process, not the process itself. For a physical system to count as "computing," a conscious agent has to first carve reality into symbols and assign them meaning. Without that agent, there are only voltage gradients. Not symbols. Not experience. Computation presupposes consciousness. It cannot produce it. The paper calls this confusion the "Abstraction Fallacy."
The analogy that makes it click: a GPU simulating photosynthesis can model every reaction perfectly. It will never produce a single molecule of glucose. Simulation is not instantiation.
The paper doesn't say artificial consciousness is impossible. It says if a system were ever conscious, it would be because of its physical constitution, not because it ran the right algorithm. No amount of scaling changes that.
This comes from inside the house. Not a philosopher. A researcher at the lab building some of the most advanced AI on the planet, arguing that the entire framework connecting computation to consciousness is logically broken.
1
59
Apr 16
Apr 16
Pete Hegseth quoted a fake Bible verse from Pulp Fiction during a Pentagon sermon.
1
1
52
Apr 13
Retardio. @ParrotCapital
Apr 13
Reporter: Did you post that picture of yourself depicted as Jesus Christ?
Trump: It wasn't a depiction. I did post it and I thought it was me as a doctor. And had to do with red cross as a red cross worker, which we support and only the fake news could come up with that one.
1
1
43
Ed Salazar retweeted
Apr 13
Trump makes a profane post threatening genocide on Easter, trashes the Pope for refusing to worship him, and posts a depiction of himself as Jesus healing the sick.
The right-wing political pastors who promoted this scumbag fake Christian as someone chosen by God are charlatans.
119
1,651
7,505
78,585
Apr 13
Da Fuk..? That the Pope was put there because "they thought that would be the best way to deal with [him]"? WTF is he smoking..? "If [he] wasn't in the WH, Leo wouldn't b in the Vatican"? "Weak on Nuclear Weapons." Wut? He's gone full retardio. @ParrotCapital @INArteCarloDoss
1
2
82
Apr 12
El Masla norteamericano. Otra piedra. @VascoAmonda
Apr 12
visits the pope → pope dies
leads Iran negotiations → talks collapse
flies to Hungary to prop up Orbán → Orbán loses in a landslide
Man’s got a streak.
2
59
Ed Salazar retweeted
Apr 10
Anthropic's Claude Mythos isn't a sentient super-hacker, it's a sales pitch — claims of 'thousands' of severe zero-days rely on just 198 manual reviews tomshardware.com/tech-indust…
99
386
2,506
424,278
Apr 10
Color me surprised... @ParrotCapital
Apr 10
🚨JUST IN: Benjamin Netanyahu just asked to postpone his corruption testimony.
The reason: national security.
1
1
55