Gökay🌒

@hhellstar_ @yiridum Use rbxts, rojo, asphalt

@haskologlu ilk lte+ ulkeye geldiginde 2014, LG G3 cihazim vardi, 450down - 200mbps upload gibi sayilari gordugumu hatirliyorum. Ulkede en buyuk sorun fiber altyapi eksikligi, baz istasyonlarin cogu sadece wireless relay yapiyor. 12 sene sonra ayni sehirde ayni konumda 5g 40mbps

Birçok bölgede 5G hız testi yaptım. İndirme hızını test ettiğim yerlerde 4G ile hemen hemen aynı. Yükleme hızında ciddi iyileşme olsa da bazı yerlerde 5G yükleme hızı 4G’den daha düşük. Sizin test ettiğiniz 5G nasıl? Ben şu ana kadar indirme hızından verim alamadım ancak Yükleme hızında iyileşme olmuş. Zamanla heryerde 5G indirme hızı artar umarım. Yaptığım aramalarda da çekme sorunu yaşanıyor. Bunla alakalı mıdır? Bilmiyorum.

@oatshaken @Laureux @LastHopebender haftada 1-2 max

@Laureux @LastHopebender yetişkin kedileriniz varsa ne kadar sıklıkla veriyorsunuz acaba

Kediyi öldürmüş. O kadar masraf ettim dediği kedinin beslenmesi hakkında veterinere hiçbir soru sormamış. Ne kadar çöp varsa hepsini vermiş 3 aylık bebeye. Allah seni ıslah etsin.

@SebAaltonen I am consuming your tweets like crazy, thank you for sharing these insights

Time to retweet my receiver shadow optimization posts from 3 years ago. Just ported the same optimization to our new WASM renderer. New engine results here (TLDR: 10M less vertices): x.com/SebAaltonen/st…

@SebAaltonen Llm wrotes code, llm wrotes test, test is full bulshit with glues adhocs everywhere, result is impressive agent think its breakthrough, and it bleed the whole context. The first thing should be done is training models that better at e2e test writing

But there will be a future where AI writes better code than human, and next gen AI will use that code as training material, etc, etc. Recursively improving itself. But we will see a several year temporary dip in average code quality before that happens.

The problem is that LLMs are writing more code than humans, but not better code. Which means that LLM learning material is getting worse, not better. Of course in the future it might go the other way, but we are not yet there today. His "one year" prediction already failed.

@NikTek Oh no but jensen said its not frame level post process 🙊

During DLSS 5 Hands-On Video, YouTube creator Hot Hardware asked Nvidia employees this question about the computational cost of DLSS 5 when you move around in the game and it really caught my eye on how they hesitated on showing any DLSS 5 footage during motion. Every time that he moved the character, he would turn off DLSS 5 first then move to a different in-game location and then turn it on, or if DLSS 5 was turned on the movement was very slow and controlled. I know this is still work in progress, but it really begs the question "why wouldn't they show any this?" or "why show DLSS 5 this early?" I tried freezing some shots when another character is in motion to check if there are any artifacts when DLSS 5 is turned on. As suspected, I did come across some artifacts with UI elements, character popping out of the UI element and ghosting/artifacts behind the character. Its still not the best way to judge these small details that you can clearly see if a footage is recorded in-game, but then again its not like Nvidia provided any of them during DLSS 5 marketing campaign. Whenever this is still going to be present by the time DLSS 5 launches remains to be seen, however as of right now it doesn't look like it will handle fast-paced motion quite well and I don't know how well this "controlled AI-generated" rendering will scale to weaker GPUs in the RTX 50 series

@SebAaltonen @kimmonismus I wonder how to tune this to your own visual identitiy, lora training? It looks bad because its generic

@kimmonismus What if the developer wants a natural looking girl without any makeup?

I seriously dont get the hate for DLSS5. Even though DLSS5 does enhance the appearance to some extent, the result is still a much more natural look. I'm well aware of the argument that this alters the developers' style, but 1) everyone is free to disable DLSS, and 2) I hardly believe that visible improvements fundamentally change the atmosphere. So far, every image has been fine by me.

@fl4ged @SebAaltonen previously only the group of experts can bootstrap these things, to much moving parts need specialized expertise on their own niche domain, languages, libraries. Now mostly you can avoid these, while you only focus on the architecture and final result both visually and performanc

@SebAaltonen and how would you know the spaghetti, so you can avoid it, surely writing a couple of renders by hand would help, so why are you proposing writing engines LLM assisted, or are you suggesting that only to the experts of the experts?

Now is the perfect time to write your own engine. The argument in the past was that you don't have time to write all the less important generic features. Now, the LLM will do that for you. You can focus on things that differentiate your product.

@ThePrimeagen Things getting worser after claude for government appeared

Claude Opus 4.6 had it's worst benchmark day yesterday

@fabio_zagami @NirmeshMehta @davidduwaer @leecronin @simocristea Read more about orch-or theory. If that's true, your consciousness is just a blink of a wave function collapse, where until quantum states collapse, it's pretty much unknown, not deterministic. The result creates an orchestration of higher structures, which looks deterministic

It is trivial to explain why a LLM can never ever be conscious or intelligent. Utterly trivial. It goes like this - LLMs have zero causal power. Zero agency. Zero internal monologue. Zero abstracting ability. Zero understanding of the world. They are tools for conscious beings.

@allfingphysics @IAI_TV I think differently, the universe is a byproduct of consciousness. Global consciousness seeks coherent patterns, reducing quantum randomness—like a glider gun in the Game of Life emerging from chaos and produces more glider guns

@IAI_TV The universe created humans to know itself better. We are searching heads for the global counciousness.

“Life developed as a vehicle for consciousness.” For decades, physicists have speculated that consciousness plays a role in collapsing the quantum wave function – the transition point where particles shift from ambiguous superpositions into definite states. But according to renowned anaesthesiologist @StuartHameroff, in collaboration with physicist Roger Penrose, consciousness isn’t just the cause of the collapse. It is the collapse itself – a fundamental process woven into the fabric of reality, with life emerging as its vessel. Watch in full here: iai.tv/video/consciou…

@wood_work16666 @StuartHameroff For one neuron, 100–500 million possible tubulin dimer in single human neuron x 6 transistors should be required to simulate these tubulin structures' oscillation correctly. It's dynamic, that's why evolution chose carbon, not silicon

@StuartHameroff 1 bit is sustained by 6 transistors. How many transistors are needed for a first approximation on a pyramidal neuron?

Thanks Sultan. Yes a single neuron can do predictive coding, especially a pyramidal neuron with the biggest arrays of mixed polarity. anti-parallel microtubules ideal for recursive processing and interference beats. Also why unicellular organisms can learn and solve traveling salesman problem. So how dumb does that make the cartoon neuron approach?! One neuron = 1 bit??! That’s what AI is basing their claims of AI inevitably being conscious on. @davidchalmers42 AI is funding cartoon neuron magical thinking. How ironic.

@daniel_mac8 Why cant just cursor etc use intelicode or similars directly

🤯 this genius stores his entire codebase syntax in a graph database and queries it so provide context to an llm

@Mascobot @a16z Ok, for a new investment for a personal AI cluster, should we wait for Nvidia DGX sparks (3000$) or build something like this?

🚨 New: We @a16z built an 8x RTX 4090 GPU AI workstation from scratch —compatible with the new RTX 5090 with PCIe 5.0, for training, deploying, and running AI models locally— so you don’t have to. Here’s how we built it, why it matters, and how you can build one too. Full guide👇

@dmipatra @kimmonismus Yes, it won't gonna improve that much. block diffusion models will arrive, millions of context sizes and hundreds of faster and lighter computation requirements. Agents will write the code, you will just be leading the direction. That's already my 1 year of workflw. I build crazy

@kimmonismus AI in current state is really stupid. They can't work with big codebases. They generate absurdic code. Even the best models. Like Sonnet. Without guiding they are useless. May be your team was not just good?

People are losing their jobs. And it will only accelerate. By the end of the year at the latest, the huge impact will be felt.

@sek_sleep @turkhafiza1 cift tarafli 2 makara var, 5km, 5km, droneda kalan kisim yaklasik 1-2 kg filan, maliyetide 100 dolarin altinda.

@turkhafiza1 Kablo dronedan pahalı olur, üstelik 1 yere girse kablo takılır, drone işlevsiz kalır..

Elektronik Harp tehdidine karşı açılan yeni sayfa, fiber optik kablolu Drone.

@iCodeSometime @RVL1389 @Dexerto client side prediction

@RVL1389 @Dexerto All the calculation should be on the server anyways

Marvel Rivals has a bug that causes players at low FPS to move slower and do less damage on certain heroes This gives players with high-end PCs an advantage

ok, my analogy here might be flawed, since it is a feedback loop, "localized chemical reactions, electrical signals, or mechanical influences" might be the main driver for subjective collapse (what else can be lol), "do microtubules somehow coordinate in these larger structures", in this sense yes, it's a loop. Higher structure effects collapse (subjective collapse)

I believe Stuart has recently discussed how simple structures can evolve without any rewarding function, this might be pointed out for bottom-up logic. A simple organism might be guided by objective reduction but Plants exploit quantum coherence (via Subjective Collapse) for energy transfer, likely sampling all possibilities to find the most optimal paths. Similarly, microorganisms might use quantum effects for pathfinding. Axons might also be influenced if microtubules exploit coherence. It’s as if until the space-time threshold is reached, or through Subjective Collapse, the system uses an effectively infinite amount of time to sample the best patterns, not just relying on localized chemical reactions, electrical signals, or mechanical influences. This brings up the idea of an equation: Axons are guided by interactions with other neurons and shaped by environmental factors (as complexity increases, their sensitivity to quantum fluctuations drops—it’s more fixed logic at that point). But microtubules amplify quantum states, stretching coherence over time and boosting the likelihood of qualia until it hits a threshold (the so-called "moment of awareness"), wave function collapse. It’s a feedback loop, constantly tweaking the probability cloud while being influenced by the surrounding environment.

Trees and plants have microtubules and are likely to be conscious at low intensity and frequency.

I am asking that myself too, but is this a binding issue? Does the quala exist before or after? Quela might drive the evolution to create structures that make quala states more coherent, so that hash was already in there, not binded afterward. Is it a bidirectional effect, a feedback loop? Too many questions

@gokayd_in @StuartHameroff so, if the "moment of self" is something like a pattern of ripples in space time, perhaps that has its own cohering binding effect. I feel like I've heard Stuart gesture in this direction but havent seen anything more concrete, hence the q