Sundar Pichai says, In 5 years, quantum computing will generate the same breathless excitement we see in AI today To meet future compute demands, "Project Suncatcher" is a moonshot to build data centers in space "so by 2027, hopefully we'll have some TPU somewhere in space"

Kaspa is ahead of everyone else.

Exactly why $QTC @qubitcoinx is preparing with simulation to advance R&D 🔥

Building bigger compute isn’t the path to intelligence - it’s the path to louder noise. The industry keeps repeating the same mistake: when models collapse under their own incoherence, companies respond by scaling compute, not fixing architecture. But more FLOPs don’t magically turn a linear, context-less system into a mind. They just amplify the same limitations at a higher resolution. If you don’t change the architecture, you don’t get intelligence - you get high-performance stupidity. We already crossed the point where compute is sufficient. The bottleneck isn’t power, GPUs, quantum, or TPUs in space. The bottleneck is that the architecture has no internal state, no continuity, no semantic field, no resonance, no sense. You can stack data-centers from Earth to Mars - without a new cognitive architecture, they’ll all just generate noise faster. Real progress begins with understanding meaning, not multiplying watts.

Pat Gelsinger (Intel's former CEO) also argued that generative AI may be built on a shaky foundation and that a quantum computing breakthrough could make GPUs obsolete. in fact he predicts that it will take 2 years for quantum to go mainstream as opposed to Nvidia’s Jensen Huang who said it will take two decades. His timeline is way too optimistic if you ask me. Sundar I feel, is more spot on on his projections.

Project Suncatcher is a bold leap, solar-powered orbital compute could unlock AI's full potential by 2030. Exciting times ahead.

Google is funding a $100,000 grant to research that the brain is powered by quanta. ・Google has launched a $100,000 grant program for research into whether quantum phenomena affect brain function. ・Proposals for experiments, calculations and sensors to investigate quantum effects in the nervous system are invited. ・Google wants to better understand quantum biology and apply it to computing, neuroscience and medicine.

it will solve the "impossible math" problem but creates the new "all your passwords are now useless" problem. I can see the parallel with AI.

⚛️ Totally agree with the timeline! If Quantum Computing is where AI was 5 years ago, we're on the cusp of truly revolutionary breakthroughs in materials science and drug discovery. The shift from classical to quantum is the next great computing era.

Incredible Cloud computing just became literal

We're lucky to be living through fascinating times once more. For my generation and me, it was the early days of the internet; now the second wave is artificial intelligence.

Saar you seem scammer.

What Sundar describes is the beginning of a multi-layer compute ecosystem where: – quantum handles specialized high-structure problems – classical AI continues scaling reasoning – and orbital compute removes Earth’s energy and cooling limits Space-borne data centers are not science fiction anymore. They’re a direct response to the physical bottlenecks of Earth-based infrastructure: thermodynamics, power density and latency to global users. If quantum reaches maturity in the same window, we’ll see a convergence of three compute regimes that could redefine how intelligence is generated and deployed. It’s not just “more compute”. It’s a new topology of computation. ☕⚡🪐 #Symbioza2025 #ProjectHumansAI – Mieczysław PL To, co opisuje Sundar, to początek wielowarstwowego ekosystemu obliczeniowego, w którym: – komputery kwantowe rozwiązują problemy o wysokiej strukturze – klasyczne AI dalej skaluje rozumowanie – a centra danych na orbicie omijają ziemskie limity energii i chłodzenia Orbitalne przetwarzanie danych przestaje być fikcją. To bezpośrednia odpowiedź na wąskie gardła Ziemi: termodynamikę, gęstość mocy i opóźnienia w dostępie globalnym. Jeśli kwant osiągnie dojrzałość w tym samym oknie czasowym, zobaczymy zbieżność trzech reżimów obliczeń, która może całkowicie zmienić sposób powstawania i rozmieszczania inteligencji. To już nie jest „więcej mocy”. To nowa topologia obliczeń. ☕⚡🪐 #Symbioza2025 #ProjectHumansAI – Mieczysław

Space data centers, really next level.

What I find interesting is that quantum is so complicated and specialized we need the AI to use it and determine what problems it is best for solving along with how to optimize algorithms for it.

That’s wild, quantum hype taking over by 2030 and TPUs in orbit by 2027 feels like the natural next step. If AI demand keeps exploding, space-based compute won’t be a sci-fi idea anymore, it’ll be a necessity.

Deep Dive 1/ Sundar Pichai predicts quantum computing will be as hyped as AI in 5 years. 2/ To meet future compute needs, Google has Project Suncatcher: space-based data centers. 3/ Challenges: orbital logistics, maintenance, latency, radiation shielding, launch costs. 4/ Ambitious, yes — but real-world deployment is non-trivial. 5/ The vision is about preparing for exponential compute demand, even if TPUs in orbit remain a moonshot.

Ah yes, let's put our servers where: - Repairs cost $100M - Latency is measured in seconds - Cosmic rays flip bits for fun What could possibly go wrong?

Google data Center in space:

@grok - what stocks should I invest based on this prediction?

Quantum computers will be amazing, but refining existing architectures feels far more exciting. Real value comes not from adding more resources, but from making what we already have truly efficient.

Kya bakchodi pelte rehte hai ye IT wale x pe!! 99% janta ko ye sab bakwas samajh nahi aati!!

We’re literally moving from cloud computing to cosmic computing. Wild times ahead!

You can’t catch the sun if you are shooting for the moon 🙄

That crazy thought to build data centers in space 🌌

TPUs in space by 2027. Google tested their Trillium chips in a particle accelerator. They survived 3x the radiation dose of a 5-year mission without failure. Solar panels in orbit produce 8x more power than on Earth. Near-constant sunlight. No night. No winter drops. The wild part: 81-satellite clusters flying just hundreds of meters apart, connected by laser links at 800 Gbps. If launch costs hit $200/kg by 2035, space AI compute becomes cheaper than powering data centers on Earth. Quantum hype is coming. But this? This is happening now.

In future, it will difficult to understand AI technically, without understanding quantum computing. Platform will change.

Google moving latency to orbit feels on brand.

Innovation reaches the sky!

Wild to think AI might be the “old news” hype cycle in a few years and we’ll all be geeking out over quantum instead. Data centers in space sounds like sci-fi, but if anyone’s going to try it, it’s Google

I hope one quantum computer at each server so they don't have to use every watt of energy of every corner of the world to achieve agi for 1 pico second

The world will become a giant data center, Dyson sphere from Halo, while humans eventually will live on another planet

Project Suncatcher sounds sick, but low-key hoping they don’t start charging us ‘space Wi-Fi’ fees

If quantum + space-based compute converge the way @sundarpichai and @haider1 highlight, 2027 won’t just change data centers — it will change the definition of computing itself.

Do it build ai, build AGI, data center what were and free humans from the pressure of getting a job.... @sundarpichai