@immanuelg
Account based inCanada
About this account
- Account based in
- Canada
- Connected via
- Canada Android App
Account-level information from X, not a live location or the device used for a specific post.
🚀 AI & Robotics Evangelist in Quebec | I help executives understand and integrate these technologies
Montreal, QC
Joined May 2009
- Tweets32.9K
- Following7.5K
- Followers4.9K
- Likes18.9K
Pinned Tweet
I did it !!!
10 billion lifetime tokens !!!
Where are all other Codex power users ? Say hi !!
Immanuel Giulea retweeted
La formation du groupe A, qui jouera au Centre Bell ce soir :
Caufield-Suzuki-Kreider
Slafkovsky-Kapanen-Demidov
Davidson-Beck-Pastujov
Thorpe-Condotta-Belzile
Hutson-Carrier
Struble-Xhekaj
Mittelstadt-Samson
Dobes
Fowler
#CH
Immanuel Giulea retweeted
La formation du groupe B, qui sera en action à Toronto ce soir :
Newhook-Evans-Berard
Bolduc-Dach-Texier
F.Xhekaj-Danault-Anderson
Poulin-McCown-Mesar
Matheson-Dobson
Engstrom-Reinbacher
Szuber-Walsh
Montembeault
Kahkonen
#CH
Immanuel Giulea retweeted
We were working on the keynote today with @romainhuet and @sama and most of the fun was trying to figure out how to explain it all to you because there is so much good stuff in there that it's a bit ridiculous all in quick succession.
We'll have some things next week already to not keep you waiting so long, but very excited to show you all new things we've been working on and how it will all come together in the coming months.
Immanuel Giulea retweeted
Just a quick reminder: this is the definition of AGI by google deepminds definition.
IMHO it’s the most scientific approach
Immanuel Giulea retweeted
For 20+ years @ShaneLegg and I've discussed AGI’s potential impact on the economy, science & society. With the DeepMind Institute, we're expanding interdisciplinary research on key questions for the AI era. We hope it spurs the discussions needed to get the next steps right: deepmind.google/institute
Immanuel Giulea retweeted
My journey to develop AGI spans 25 yrs, including 10+ yrs thinking about technical & societal perspectives at Google DeepMind.
AGI is on the horizon - we need deeper understanding of its implications. To help, we've created the DeepMind Institute.
Nvidia CEO Jensen Huang just said:
"I expect Nvidia to sell twice as many chips this next year as we do this year. And the reason for that is because AI has has so much contribution to the benefits of different industries, different economies. And you could see that almost every single country that we're in, people want to invest in AI."
Nvidia does not (has not) given unit growth commentary like this I believe? The street sees revenue of $410B in fiscal 27 (consensus). $216B in fiscal 26.
He's likely talking beyond the GPU, right? He might be referring to everything including Vera CPUs, Groq LPU
Welcome thoughts, ideas and attention to this -- thanks
Immanuel Giulea retweeted
We are releasing the latest version of BrokenArXiv and ArXivMath! These benchmarks now focus on conjectures that were refuted in the last month on ArXiv, and models are executed within a harness instead of directly via API.
Performance remains impressive, with GPT-6 Astra on top
Immanuel Giulea retweeted
We just announced Canada’s new Productivity Mega Deduction.
This cuts Canada’s marginal effective tax rate from 13% to 6.4% on new business investment — the lowest of any major economy in the world.
Canada is building big — and becoming the best place in the world to build.
More here: pm.gc.ca/en/news/news-releas…
SpaceX is currently targeting Starship Flight 14 for September 22 🚀
And this flight is a huge step beyond everything Starship has done so far
Flight 14 will attempt to:
• Reach stable Low Earth Orbit for the first time
• Deploy production Starlink V3 satellites into orbit
• Become Starship’s first revenue-generating mission
• Attempt the full orbital mission profile from launch → deployment → deorbit → reentry
• Push the V3 heat shield through another major test
• Continue proving Raptor 3, guidance, landing burns and vehicle control needed for rapid reuse
The hardware:
• Super Heavy Booster 21
• Starship Ship 41
• Third flight of the Starship V3 architecture
• 33 Raptor 3 engines on Super Heavy
• 6 Raptor engines on Starship
• Launching from Pad 2 at Starbase
After liftoff, Super Heavy will separate from Starship, perform its boostback and return for a controlled ocean splashdown
No tower catch is planned for Flight 14
The really big part happens with Ship 41
Instead of following another suborbital trajectory, Starship will continue into orbit
Then it will deploy production Starlink V3 satellites into stable LEO
That is a massive milestone by itself
Flight 13 proved Starship could deploy V3 satellites in space....Flight 14 is the first time Starship is expected to actually leave production V3 satellites in orbit
And these satellites are monsters
Each Starlink V3 is designed for:
• 1 Tbps of downlink capacity
• 160 Gbps of uplink capacity
• 2,048 beams in each direction
• Six 400 Gbps laser links
Eventually a single Starship will be able to carry up to 60 V3 satellites....adding around 60 Tbps of Starlink capacity in ONE launch
Around 20× the downlink capacity deployed by a Falcon 9 Starlink launch
After payload deployment, Ship 41 will perform its deorbit sequence and come back through the atmosphere
That means another major test of:
• The orbital heat shield
• Flap control through reentry
• Guidance during the return
• Raptor relight
• Landing flip and burn
Ship 41 will then return for a controlled ocean splashdown
And the next step is already lined up....Elon has confirmed Flight 15 is planned for the first attempt to bring Starship all the way back to the tower and catch it
But Flight 14 comes first
Reach orbit....Deploy real Starlink V3 satellites.....Come back through reentry
This is where things start getting very real
Immanuel Giulea retweeted
This is potentially a great news for robot learning:
Odyssey-3 (a foundation world model) just dropped and its attacking robotics' biggest bottleneck, task-specific data.
Odyssey's approach is to start with a pretrained world model, then attach an action decoder trained on tens of hours of robot demonstrations.
Odyssey-3 drove real Indian streets after its driving policy trained on only 20 hours of simulation.
A lot of physical AI progress still comes from throwing more demonstrations at one narrowly defined system. Odyssey is arguing that broad world-model pretraining can absorb much of the knowledge that every downstream policy would otherwise need to relearn.
Their reported numbers are early but unusually small: tens of hours for robot arms, tens of hours of humanoid teleoperation, 20 hours of simulated driving, and tens of hours of simulated drone data.
The adaptation layer then learns how to turn Odyssey-3's representations into the actions of that specific system.
This is basically a data-allocation bet. Spend the huge pretraining budget once on a general model of the world, then spend far less experiential data teaching each new body its controls.
🤝 Paid partnership
Immanuel Giulea retweeted
2026 is the year of efficiency and Tuesday is for random swag drops on your door handle
Immanuel Giulea retweeted
Jensen Huang completely destroys Jacob Coxon’s (ex-Anthrpic employee) 10% extinction prediction. no data, no scientific grounding, no defensible 10%.
"First of all, we shouldn’t [try to explain a 10% chance of extinction], because it’s made up.
These are well-educated people. They’re called researchers. Obviously, they’re working in a lab, and so the confluence of these words—and then the prediction—is alarming and troubling.
It shouldn’t be done. It’s irresponsible."
----
Full video on "All-In Podcast" YouTube channel (link in comment)
MUST SEE: Amazing moment as President Trump calls Nvidia CEO Jensen Huang while he's on stage at the All-In Summit.
@POTUS on AI Doomerism: “I'm telling you, it's all a hoax… and we're not going to let that happen.”
Immanuel Giulea retweeted
Fantastic news!
This will draw $ billions of investment in to our major airports, while radically improving the passenger experience.
Just visit some of the major airports in Asia and the Middle East to imagine what is now possible here in Canada!
Immanuel Giulea retweeted
This could be the game changer that, together with other recent policy changes and tightness in global energy markets, finally unlocks tens of $ billions in new green field oil sands projects! 👏👏👏
At the very least, it will be a huge help to unlocking major new capital investment in Canada.
Prime Minister Mark Carney just announced a “productivity mega deduction” that allows companies to immediately deduct 100 per cent of the cost of new investment across far more assets than previously allowed.
It makes a tax measure announced in last year’s federal budget, the “productivity super-deduction,” more generous.
The new regime roughly cuts Canada’s marginal effective tax rate on new business investment in half, from 13 per cent to 6.4 per cent. The government says that is the lowest rate of any major advanced economy.
🔗 Follow our reporters' live coverage of all-things related to the investment summit on Day 2: theglobeandmail.com/business…
Immanuel Giulea retweeted
Poilievre boude.
Le seul chef conservateur du XXIe siècle à avoir remporté trois élections générales participe au sommet de Mark Carney les 14 et 15 septembre; sa présence apporte une crédibilité conservatrice et envoie aux investisseurs un message de continuité et de stabilité.
D’anciens ministres comme James Moore et Jason Kenney saluent sa participation, tout comme des libéraux qui ont jadis combattu son gouvernement.
Pourtant, à ma connaissance, Pierre Poilievre et tout son caucus gardent le silence sur Harper, à l’exception de Gérard Deltell.
Le pays passe avant le parti. On peut demander des résultats à Carney sans bouder la contribution de Harper à une initiative qui pourrait servir le Canada.