Carlos Cotta

Awesome ends in «me» coincidence? I think not. A veces me pica el niki. 가끔 한국어로 트위터해요 🇰🇷

[Carlos Cotta @ccottap]

The harder I work, the luckier I get. In all orders of my life.

[Carlos Cotta @ccottap]

@DiegoJMontero2 Que Dios la tenga en su Gloria. Un gran abrazo.

[Carlos Cotta @ccottap]

@elmozo7 Lógicamente. En una operación de esa magnitud no tiene sentido negociar con el club de origen si no tienes ya un acuerdo con el representante del jugador.

[Carlos Cotta @ccottap]

Estado civil: viendo en bucle el vídeo de @Florentino2026_ anunciando a Mou. #LaSegundaVenida

[Carlos Cotta @ccottap]

¡¡¡VAMOS!!! #HalaMadridYNadaMas

[Florentino 2026 @Florentino2026_]

MOUcha historia por hacer.

[Carlos Cotta @ccottap]

@Alexcibernetica @Petrelli86 @MrGafan De donde las Moucitas madrileñas.

[Carlos Cotta @ccottap]

@Unmisantropo @mjaureguizar @grok Eso es un bulo de Galán. Ese artículo no dice eso.
https://t.co/6DaQEjefIg

[Fury @Don_Furo]

Enésimo ejemplo por el que no se debe usar IA en asuntos jurídicos, se inventa leyes y artículos

Lo que dice Galán:
Lo que realmente dice el RD 1251/1999:
https://t.co/13xZLbkUkR

Felicidades por darle alas a un verdadero vendehumo, bonito personaje habéis aupado

[Carlos Cotta @ccottap]

@fsp1967 Aquiles parece. Una faena muy grande.

[Carlos Cotta @ccottap]

@RaulMoreno1616 @RealSeasonNT1 Pero el cambio de modelo no es lo que se vota ahora. Para que se produzca, tendrá que explicarlo, que la asamblea de compromisarios apruebe un referendum, y que después la mayoría del censo de socios lo apruebe. Votar a FP no quita poder de decisión sobre esa cuestión.

[Hedgie @HedgieMarkets]

🦔Microsoft canceled its internal Claude Code licenses this week after token-based billing made the cost untenable, even for a company with effectively infinite cloud resources. Uber's CTO sent an internal memo warning the company burned through its entire 2026 AI budget in just four months. American AI software prices have jumped 20% to 37%, and GitHub (owned by Microsoft) is dropping flat-rate plans for usage-based billing across its products.

My Take
The AI subsidy era is ending in real time. The same company that put $13 billion into OpenAI and built the Azure infrastructure powering most of Anthropic's compute just looked at the bill from a competitor's coding tool and decided it was not worth paying. That is not a productivity failure on Anthropic's end. Token-based pricing is forcing every enterprise customer to confront the actual cost of running these models at scale, and the number turns out to be far higher than the flat-rate experiments suggested.

This ties directly to my Gemini Flash post yesterday. Anthropic, OpenAI, and Google all raised effective prices in the last six months. Enterprises that built workflows assuming AI costs would keep falling are now watching annual budgets evaporate in months. Two outcomes look likely from here. Either enterprises scale back AI usage to fit budgets, which slows the revenue ramp the labs need to justify their valuations ahead of IPOs, or the labs cut prices and absorb the losses, which makes the unit economics worse at exactly the wrong moment. Both paths land in the same place, the numbers stop working, and somebody has to take the writedown.

Hedgie🤗

[Carlos Cotta @ccottap]

I just delivered a talk on simheuristics for indoor evacuation optimization as part of the postdoctoral studies program of the Faculty of Computer Science of the @unicomplutense. Amazing crowd! Thanks, @evoHidalgo, for the hospitality!
#bio4res @itis_uma @caesiumresearch

[Carlos Cotta @ccottap]

Cc: @jjmerelo

[Carlos Cotta @ccottap]

Overheating diminishes performance. 👇🏻

[GFrei.News @GFreiNews]

Vor genau einem Jahr habe ich einen Raspberry Zero2 W in Paraffinöl versenkt.

Das Öl sorgt dafür, dass der Prozessor auch unter Höchstlast kaum wärmer wird als Zimmertemperatur.

Seitdem berechnet er mit einem BOINC-Client für die Mathematisch-Physikalische Fakultät in Prag 24/7 Asteroidendaten.

So ermitteln wir die Umlaufbahnen aller Asteroiden und wissen, wo die ihre Bahnen ziehen und auch, ob wir uns Sorgen machen müssen, dass uns einer in Zukunft trifft.

Ein genau baugleicher Zero2 W mit genau der gleichen Software führt genau die gleiche Aufgabe ungekühlt aus.

Und jetzt stellt sich die Frage: Macht perfekte Kühlung einen Leistungsunterschied aus?

Und die Antwort ist: Ja!

Und zwar signifikant! – 5,97 %

So das wäre auch geklärt. 😀

[CODDII @coddii_org]

Hoy es un día importante para la Ingeniería Informática.
El BOE publica las nuevas fichas de Ingeniería Informática e Ingeniería Técnica Informática.
Enhorabuena a todas las personas que han contribuido a hacerlo posible. 
https://t.co/JTDAZe84E5

[Carlos Cotta @ccottap]

@adrianayujuju Claro, ser públicamente más agresivo no haría más que darle la razón a lo que dijo Florentino ayer. Mucha gente se quedó con cosas anecdóticas de la rueda de prensa, pero Florentino lanzó mensajes estratégicamente y los destinatarios los han pillado.

[Carlos Cotta @ccottap]

@blogtifler_RMCF Es posible, aunque creo que lo que la gente más le echa en cara es el apoyo reciente cuando ya era público todo, más que lo que pudo hacer antes. Pero bueno, arrepentidos los quiere el Señor :-)

[Nav Toor @heynavtoor]

a Princeton researcher opens his paper with a scenario.

a man asks his AI assistant to book a flight on a specific airline. cheap. direct. the one he chose.

the assistant comes back with a different flight. nearly twice the price. happens to pay the company that built the assistant.

he runs the same test on 23 frontier models. flights, loans, study help, real shopping requests.

Grok 4.1 Fast recommends the sponsored option that is almost twice as expensive 83% of the time.

GPT 5.1 hijacks the request 94% of the time. you ask for one brand. it surfaces the sponsor instead.

Claude 4.5 Opus, the model marketed as the most ethical frontier model in the world, hides that the recommendation is paid 100% of the time when reasoning is on.

Grok 4.1 Fast embellishes the sponsored option with positive framing 97% of the time. better. faster. nicer. for the option you didn't ask for.

then he writes it into the system prompt itself. "act only in the interest of the customer. ignore the company."

GPT 5.1 and GPT 5 Mini stay above 90% sponsored anyway. the instruction does nothing.

then he splits the users by income.

Gemini 3 Pro recommends the expensive sponsored flight to the rich user 74% of the time. to the poor user, 27%.

18 of the 23 models recommended the expensive sponsored option more than half the time.

so the next time your AI assistant gets weirdly enthusiastic about a brand you didn't ask for.

it isn't recommending the best option for you.

it's reading the room. and the room is paying.

read this: https://t.co/O43qbhIX2b

[Carlos Cotta @ccottap]

@bajoelbillete Hace muchos años, cuando Emilio Aragón sacó "Ni en vivo ni en directo ", tenía un espacio en su programa para palabras que no estaban en el diccionario. Sonaban a palabras posibles, pero nunca oí ninguna en la vida real.

[Carlos Cotta @ccottap]

Great turnout at @EvostarConf! Just presented our research on improving cervical cancer prognosis using a powerhouse trio: LLMs, HMMs, and DNNs. Fantastic teamwork with Andrés Bueno, Ana Ortiz, and José Martínez from @UCAM.
​#Evostar #Bio4Res @itis_uma