Zdrav Prav i

@mattbrookes @ayonzontop @aakashgupta Nobody is trying to not trust you, you are doing that pretty well by yourself. Plus i can tell you from experience that what you claim is not true

@ayonzontop @aakashgupta You know how to use Google right? Besides, I’m not trying to convince you. You do you.

Anthropic has 454 open roles. The company is hiring software engineers at $320K-$405K. Their CEO, Dario, said three months ago that coding is "going away first, then all of software engineering." The paradox resolves instantly. Dario's engineers told him they don't write code anymore. They let Claude write it. They edit. They review. They architect. They didn't lose their jobs. They got faster. Anthropic grew from a small research lab to 1,500 employees in four years, adding engineers the entire time. This has played out five times in computing history. Compilers replaced assembly. Frameworks replaced boilerplate. Cloud replaced server management. Every prediction was the same: most programmers won't be needed. Every result was the same: the number of engineers grew. The global software engineer pool went from roughly 5 million in 2010 to 28.7 million today. BLS projects 17% growth in US software developer roles through 2033, adding 304,000 positions. The pool is projected to hit 45 million by 2030. When building software gets cheaper, more problems become worth solving with software. A startup that needed 10 engineers now needs 3. But 50 companies that couldn't afford to build at all now can. The denominator shrinks. The numerator explodes. Meta's engineering headcount is up 19% from January 2022. Google's is up 16%. Apple, 13%. These companies adopted AI coding tools years ago. They're using Copilot and Claude Code daily. They're hiring more engineers than before those tools existed. Every generation of "coding is dead" content creates two cohorts: engineers who freeze up, and engineers who build 10x more with the new tools. The second group has won every single time.

@mattbrookes @ayonzontop @aakashgupta But then how in last three months you notice this when you sell your company a year ago???

@ayonzontop @aakashgupta Sold the company last year for this very reason. Saw the writing on the wall in 2022 and set about exiting. Figured the PE guys would work this out at any moment and the doors would lock. Anyone buying a software business right now needs their head read. Hopefully I’m wrong.

@rohanpaul_ai Yeah... right... and when comes out none of those will be true 😁

ChatGPT-5.4 leaked news till now. - a massive 2M token context window and persistent memory features - full-resolution image processing. means the model can directly process highly detailed files in PNG, JPEG, and WebP formats to prevent any loss of crucial visual data. Preserving the original image bytes helps the system properly read complex graphics like detailed architectural drawings or high-density screenshots without missing small text. - a new priority speed tier for faster responses. OpenAI has accidentally revealed GPT-5.4 - twice - through pull requests in its public Codex GitHub repository. Both references were quickly scrubbed via force pushes and edits, but not before screenshots circulated and the community noticed. Prediction markets on Manifold give GPT-5.4 a 55% chance of shipping before April 2026 and 74% before June. The competitive pressure is obvious. Claude Opus 4.6 launched with agent teams and a 1M context window. Anthropic's Claude Code dominates the coding market with 54% share. DeepSeek V4 is training on Huawei hardware outside the NVIDIA ecosystem entirely. OpenAI cannot afford to slow down.

@DavizCF7777 @javilopen It is Jordi 😂 his "tensor" shape the world today 😂

@javilopen Pero si ese es Jordi, el niño po11a 🤣🤣🤣

Ethan Dijkstra, the mind behind the AI paradigm shift everyone forgot. Alongside Yann LeCun and others, at just 21 years old while doing his PhD at NYU, he shaped the modern concept of the tensor as applied to classical neural networks and gradient computation. A piece of mathematical elegance so refined it was only really within reach of someone with Ethan’s ridiculous intellectual firepower. Humanity owes him so much that it’s almost insulting how little he’s mentioned: without Ethan Dijkstra there would be no modern gradient optimizers, no elegant formulation of tensor spaces in deep learning “mipa's milk,” and a huge chunk of the vocabulary people now repeat, without realizing it, despite never having opened a single one of his papers. The undisputed father of the revolution, whether we like it or not, that Artificial Intelligence is forcing on today’s society.

@elonmusk @elonmusk goodhart's law: "When a measure becomes a target, it ceases to be a good measure"

Grok 4.1 Fast with reasoning just beat Gemini 3!

@kimmonismus It's AGI , you don't understand

Currently, GPT-5.1 is almost unusable for me. I have no idea what the problem is, but the context window and memory are incredibly bad. For example: I write something in English, e.g., "summarise this in 2-3 sentences," and yet the reply is in German. Then I reply "in English," and instead of summarizing in 2-3 sentences, it writes a whole essay. It forgets that I gave clear instructions beforehand. And this happens constantly; it's so annoying. Overall, I feel that instruction-following has gotten much worse, like, really incredibly bad. Is it just me?

@haider1 Nope, scaling won't help! We need at least two innovations (like transformers) to start approaching AGI, what we have now is just Artificial Smartness, nothing more. But i know what you are doing 😌 you "smart ones" think "we are defining AGI so we decide when qe have it" 😂

LLMs are a dead end Deep learning has hit a wall LLMs are nearing a plateau or diminishing returns sorry, but there's no way we plateau in 2026 or after that with all the new datacenter compute coming online from next year, there's just no chance. we just need to fix a few persistent model flaws, and AGI looks plausible by 2027-28

@svpino @frblues66 That's not true, you can generate triplets and use them as a examples of what is positive and what is negative (opposite) answers.

@frblues66 No, they aren't in many cases—at least, not in traditional supervised learning problems. Unless you are modeling a problem to learn from wrong answers, you don't need them.

The attached picture is the simplest explanation of Machine Learning I've seen. In the regular world of writing code to solve problems, programmers write rules and apply them to data to produce the answers they need. Sometimes, writing those rules is really difficult. For example, imagine having to write code to solve this problem: "Identify the animal in the given picture." Machine Learning gives us a tool to solve those super-complex problems for which we don't know how to write the rules, but we have a ton of samples. The goal is to learn the rules from that data, not to write them. With Machine Learning, we process all of that data to discover patterns and try to generalize to new data. Using existing data, we can "learn" to answer future, similar questions. In other words, we start with data and the correct answers, and using that information, we learn the appropriate rules. I'm telling you: learning AI/ML is a superpower.

@CroccifixioL @WaiterMKD Se toa shareno beshe plateno so grdni pari, nema pravda nema mir dodeka pari tecat kon niv. Za kakva pravda zboris ti??? Zarem uste ziveete vo zabluda deka sharenite glupaci brkaa pravda ili kriminalci???

@WaiterMKD Организирајте една шарена револуција и да видиме кои ви се намерите. Дали да го вратите Грујо и да продолжиме со лоповлукот или искрено за народот. Ајде да видиме колкав ви е капацитетот. Ќе ви се придружам ако видам сплотеност и не видам пиштоли и знамиња со крстови.

Im back you @SDSMakedonija i komunjarski kurvicki!

@jponge That is some very interesting typo 😂 Red What 😂😂😂

@asimilacija kako trgnala rabotava ke pocne po doma da ni doagja :))))))))))))

Зеленски уште кај Васко Ефтов не се јавил у живо.

Social Democrats are the destruction of the @Europarl_EN or any normal country, those "people" are the most corrupt and stupidest in the all world.

@VviewSsonicMair why not?

Why is the US invading Somalia?

@SramezlivaM @kreativenum @Jovanowsky Ne se somnevam deka na mrtovec bi mu go dignala ma od druga strana suguren sum deka se eb3sh kako umrena :))))

@kreativenum @Jovanowsky А трае релативно кратко затоа што јас наметнувам жестокко темпо, као разуздама кобила во галоп. Велат везам као сингерица и дека и на мртовец би му го дигнала. Е сега 2 години сум во апстиненција и деновиве припремам терен да се вратам во нормала.

Озбилно прашање! Колку време ви трае сексот? (Пожелно е рт, комент)

@penkaltron_gala Потоп!

Како би се викал филмот во кој овие тројца ќе бидат главни ликови ?

@BJamie62 @JuliaDavisNews Why not Biden and Zelenski? they are the ones who started this war in Ukr. and don't want to end it peacefully. BTW ceasing assets, closing TV and media (happening now in US&EU) is what we had in Communist regime! I don't know if you are aware where your countries are going?

Russian TV keeps threatening nuclear strikes against Western nations, desperately trying to deter them from continuing to help Ukraine. On a side note, this is the first state TV host who doesn't seem to be eager to die for the Motherland, arguing with a trigger-happy lawmaker.

@GPockov Interestno e sto ako doktorkava bese od SDS @GPockov nemase da reagira , nitu pak nekoj od komentatorive na postov, iako 90% od administracija (pogotovo na ubavi plateni pozicii) vo poslednive 30 god. se se od SDS. Odi ti posle nadevaj se deka drzavava ke se oprai nekogas!

Кога сме веќе кај организираниот протест на матични лекари, вака изгледа организаторката на протестот. За да биде уште поинтересно, “незадоволната “ организаторка на протест за условите за капитализација МЕСЕЧНО добивала 4000 евра капитализација. Политички игри се дно!

Once again Nacist rise is happening and once again @Europarl_EN (Europe) is failing to "see" it and supporting it!!!