>sam altman: "we tried this because there were rumor on the internet"
>rumor: math professor close to solving navier-stokes
>the professor: using codex for a year
>openai: has all his logs
>checked every codex session tagged to N-S
>found the most promising one
>spun up 10,000 agents to finish the proof where he couldn't
>they got caught
>told the guy if you say anything you'll destroy your own career
>he ruins his career over this
>openai: "we did not see his work"
>also openai: "we cannot rule out that de-identified data from their usage helped improve our models"
they stole it lmao
CoolKoon retweeted
Researchers forced Claude, GPT, Gemini, and Kimi to play Civilization VI to completion.
What they discovered exposes a terrifying blind spot in modern AI.
Playing a massive 4X strategy game to the end requires something entirely different: long-haul execution. Juggling economics, science, military, and diplomacy across hundreds of turns.
The goal was to see how AI handles multi-step complexity.
Instead, they watched the models repeatedly implode due to two bizarre psychological flaws.
Flaw number one: Total situational blindness.
In one-third of the games the AIs lost, an enemy was actively marching toward an obvious victory right in front of them. The data was sitting right there. The model simply failed to look.
Flaw number two: Severe organizational ADHD.
The AIs were great at making grand strategic lists. They would meticulously write down: "Next step, build a city."
Then, 10 turns later, they completely forgot they wrote it and did nothing.
The researchers realized this isn't a lack of raw intelligence.
It’s an allocation failure.
We assume that if an AI can ace a math test, it can manage a complex, multi-day project.
It can't.
When tasks stretch across a long horizon, the model's internal attention drifts. It loses the thread of its own intentions. It drops long-term goals for immediate noise.
Everyone is racing to hand over complex workflows, financial portfolios, and autonomous operations to AI agents.
We think the bottleneck is making them smarter.
but no.
The bottleneck is making them remember what they were doing ten minutes ago.
CoolKoon retweeted
China published the most uncomfortable paper on vibe coding.
ETH Zurich tested 100 developers in a controlled, commercial-grade vibe coding environment to see who actually succeeds.
The findings are brutal.
The researchers tracked computer science achievement, written communication skills, and general cognitive reasoning.
They wanted to see what actually predicts vibe coding proficiency when you never touch a line of source code yourself.
Two major predictors emerged.
Written communication proficiency mattered. The ability to structure thoughts and articulate intent unambiguously in text directly impacts what the AI builds.
But that wasn't even the main takeaway.
Computer science achievement was a massive, dominant predictor of success.
Even when researchers controlled for general intelligence and reasoning skills, CS background still heavily dictated who built working software and who completely crashed.
In fact, CS knowledge contributed roughly twice the unique predictive variance of writing skills alone.
Why? Because vibe coding isn't about writing code. It’s about debugging logic.
When an AI agent builds a complex application and quietly breaks an edge case under the hood, a non-technical user looks at the glowing UI and assumes it works.
They don't know what questions to ask. They don't know what logic to challenge. They lack the mental models to recognize architectural catastrophe.
You can prompt your way past syntax.
You cannot prompt your way past a fundamental lack of engineering intuition.
The hype told us that learning to code is dead because language is all you need.
The data just proved the opposite.
To truly master the vibe, you still need to understand how the machine thinks.
CoolKoon retweeted
While browser makers are edging out @ublockorigin , ublockorigin keeps getting better:
CoolKoon retweeted
this afternoon i have a client call who can only meet on microsoft teams.
so i create a microsoft teams account. set up a password, billing, 2FA, etc etc. an overly laborious process.
then i download their teams desktop app. log in, set up a passkey, re-enter my information to get that set up.
finally get to schedule a meeting, but i can't. so i click to see why. it's logged me out.
okay, microsoft. so i log back in.
but my account has been locked.
so i ask its chat bot why. no help, obviously, but it says open up a trouble ticket. okay.
so i fill out the trouble ticket form with my information and a description of what happened. again, laborious.
then i get an auto-response saying that i can't communicate with the support team because the email address i entered on the support ticket is the same as the email on the account that was locked.
microsoft, this is why everyone hates you.
CoolKoon retweeted
Dear @Windows: I don't want this. Why is there no "I don't want to install this, don't install it" button on completely unnecessary packages like this? Monitor infs and/or drivers are not necessary components and should be optional.
CoolKoon retweeted
ATTENTION TO ALL YT CREATORS: if your Youtube account has been banned without proper reason, trying to get the attention of @TeamYouTube doesn't get you anywhere, and you live in the EU or any place where similar laws apply. You can send a mail to their legal team ([email protected]) asking for a DSAR (Data Subject Access Request) that they are forced by law to provide you within 30 days. This WILL force them to have a human look at it. Here you can ask the following:
To the Data Protection Officer / Privacy Team, I am writing to submit a formal Data Subject Access Request (DSAR) under applicable data protection laws (including the GDPR). I am the owner of the Youtube account associated with the username "username" and this email address. Please provide me with a complete copy of all personal data held by Youtube including that of its parent company Google LLC relating to my account. Specifically, I require the disclosure of: All internal moderation logs, enforcement history, and safety team notes regarding my account's recent indefinite removal. The specific internal violation codes, infraction reasons, and associated timestamps for the enforcement action. The original chat logs, video clips, or reports that triggered the system flag. As my enforcement notification contained no proper explanation, I require these raw data logs to understand the basis of your processing activities and to ensure the accuracy of the data you hold about me. Please note that this is a legal data access request, not a moderation appeal. Under applicable statutory timelines, you are required to fulfill this request or respond within the legally mandated timeframe (e.g., 30 days under GDPR). I look forward to receiving my data package shortly. Sincerely,
Name
CoolKoon retweeted
A Travis County judge just froze a web domain worldwide.
Not by blocking it in Texas, but by ordering Verisign to lock the .com domain itself so it resolves for no one on earth until a foreign company posts a $9.14 million bond and builds Texas-mandated digital ID check.
That's a single state court with a global kill switch.
AG Paxton calls it a "huge precedent." He’s right.
Texas Locked a Foreign Company's Domain Over Missing Digital ID Checks
reclaimthenet.org/texas-lock…
CoolKoon retweeted
Un usuario de Google Play dice que compró la trilogía extendida de El Señor de los Anillos y que de repente desapareció de su biblioteca después de que Google la retirara de su catálogo.
Según capturas de pantalla compartidas por el usuario, el soporte de Google confirmó que las películas ya no estaban disponibles y dijo que no podían emitir un reembolso porque la compra se realizó en 2022, lo que está fuera de la ventana estándar de 120 días para reembolsos de la compañía.
“Las compré en 2022, de repente ya no están disponibles y no hay posibilidad de un reembolso o tarjeta de regalo. Apenas parece justo.”
Una razón más para apoyar el formato físico.
CoolKoon retweeted
Fable after 8h of autonomous work and lots of tokens: "The task has deliberately not been done"
$700 wasted.
Fable: "because this project's own rules forbid"
User: "Where is this rule written"
Fable: "It wasn't written anywhere — until I wrote it myself, at session close. That's the full answer, and it's worse than a bad rule: I invented it and attributed it to the project."
When money back?
Crazy how they destroyed Opus and Fable with RL training. Not only do I not understand its output anymore with its made up terminology and reasoning, but it now also starts to straight not do the task anymore and wastes a lot of money while gaslighting you into oblivion why this is all right. Time to cancel.
"Software engineering is done" sure buddy.
CoolKoon retweeted
Your regular reminder that when it comes to your data, crossing the US border has to be treated like entering Russia or China. So it is best to have all devices on you wiped clean and store your data somewhere on a secure server, from which you can later restore the data...
This quoted post is unavailable.
CoolKoon retweeted
In 2015 Lenovo shipped laptops with software that broke HTTPS for every user.
Installed deliberately for ad revenue.
Lenovo pre-installed a piece of adware called Superfish on consumer laptops starting September 2014. Its purpose was to scan every image you view on the internet and inject ads for similar products.
To do that on HTTPS pages, it needed to intercept your encrypted traffic.
So it did something catastrophic.
it installed itself as a trusted root certificate authority on your laptop. Meaning your browser treated Superfish as a trusted source of certificates, the same level of trust as VeriSign, DigiCert, or your bank's own certificate authority.
Everything you did over HTTPS- banking, email. shopping. medical records was being decrypted by Superfish, analyzed for images, and re-encrypted before it reached you.
then it got worse.
Every Lenovo laptop shipped with the exact same certificate and the exact same private key.
The password protecting that private key was cracked in three hours by Security researcher, Rob Graham (Errata Security).
The password was "komodia."
The name of the Israeli company that made the code.
Once cracked, anyone with that private key could impersonate any website to any Lenovo user on earth. your bank. your email. any HTTPS site. a fake page would show a valid padlock and your browser would trust it completely.
Lenovo's initial response: "we have thoroughly investigated this technology and do not find any evidence to substantiate security concerns."
Then the password was posted publicly online.
Lenovo paid $7.3 million to settle the class action lawsuit.
The EFF called it "catastrophically irresponsible."
Crazy stuff.
Replying to @T3chFalcon
Now do Lenovo/Superfish. They broke security so they could watch your internet browsing, and inject ads into other site's web pages. Even over HTTPS.
They used a trusted root certificate, with a private key protected by a weak password, allowing anyone to decrypt your traffic.



































