Rob Imbeault

Effort is the algorithm

Whatever happens to the feed let’s keep it!

@martin_casado not sure why I liked this but...like

VAMOOOOOOOOSSS!!!!!!!!!!!!!!!!!!

@NatureUnedited can confirm

Studies show that watching a beaver eat cabbage lowers stress by 17%

Buying more GPUs or switching providers isn't always the answer. The bigger opportunity is getting more performance from the models you already run. That's why we built BackboardQuant. BBQ (yes, we call it that) is designed to compress models without sacrificing the performance that makes them useful. If bringing AI in-house is part of your roadmap, start with the benchmarks. 👉 backboard.io/research/quant…

@melbourneandrew 🙌

The most important work we do at 8090 is not building Software Factory. It is using our systems and applying our techniques to improving the rules-engines that make our civilization functional.

@Jason @markgurman Two 80B Qwen models operating concurrently 100% local on a MacBook Pro. I totally agree. I think Apple is gonna pop in the next 18 months because we can do so much locally.

I really want to buy more Apple given the new CEO being a hardware guy AND @markgurman report that they will release M6/M7s with absurd amounts of RAM... Stock price is fully valued, but when it dips I'm gonna buy more. ... I would have no problem buying every team member @launch a $50,000 desktop computer with 1TB or memory. I can't think of better use of a $1M, than giving each associate at the firm a desktop capable of running a FABLE level model.

GLM 5.2 was a bigger leap in model performance than Fable. What makes Fable so great is the tool orchestration around it. You can think of it as harness+

In 1899, Wilbur Wright wrote a letter to the Smithsonian, asking for all the materials they had on the study of flight. He and his brother had no experience other than building a makeshift printing press and building bicycles. The Smithsonian sent them materials. The brothers studied the birds and created models and iterated. This all in the face of public mockery and inventors before them plummeting to their death. You can just do things.

@venturetwins Congratulations 🎉

Since this is Instagram now, some personal news 💍 (he’s not online so I’m protecting the identity of the innocent 😂)

Mistakes are a privilege of the active

@rohanpaul_ai So important that the hardware that can support local frontier-level LLM inference is coming and with lossless compression/quant and other techniques everyone can have their own personal AI.

Apple is reportedly building an M7 Ultra chip that supports up to 1.5TB of unified memory. That would more than double the memory ceiling on today's M3 Ultra Macs. The push comes from local AI, where large models need vast memory to run. The big deal is that M7 Ultra’s GPU could access the entire 1.5TB pool directly. Unlike desktop DIMMs, Apple’s unified memory is package-integrated and fixed when owners purchase it. An Nvidia Blackwell B200 server GPU carries 180GB HBM3e and reaches roughly 8TB/s bandwidth. 8 B200 GPUs together provide 1.44TB, so if M7 Ultra really happens then it will make it genuinely server-scale. Unified memory lets the CPU, GPU, and Neural Engine share one fast pool. So data moves with lower latency and less power than split PC memory. i.e. Apple’s memory would serve CPU and GPU together, unlike NVIDIA’s dedicated GPU-only HBM. Apple’s shared pool reduces copying between processors and gives the GPU far more capacity. Apple's M3 Ultra already reaches 819GB/s of memory bandwidth by fusing two Max dies. Bigger memory would let you run larger models and datasets entirely on-device. That pulls Apple's top Macs into enterprise workstation and AI server territory. It moves closer to dedicated accelerators like Nvidia's Blackwell, without claiming parity. The catch is supply, because DRAM prices are climbing and parts stay scarce. Apple already pulled its 128GB Mac Studio this year amid those shortages. An a 1.5TB part would need far more of that same costly memory. Apple has not confirmed the M7 Ultra, so its final specs could change. Memory supply, not chip design, decides whether this ships at a sane price. If DRAM stays scarce, only film studios and labs could afford a 1.5TB Mac. For some comparsion, an Nvidia RTX 5090 carries 32GB VRAM, but delivers 1.792TB/s bandwidth for graphics workloads. M3 Ultra reaches 819GB/s, so Apple currently offers more capacity but less bandwidth. --- voice .lapaas.com/apples-m7-ultra-chip-may-support-massive-1-5tb-ram-capacity/

@levelsio I want to go to Oslo

🏆 Places the most attractive, rich & educated visit 1. 🇺🇸 Boston 2. 🇨🇭 Zurich 3. 🇿🇦 Cape Town 4. 🇨🇭 Geneva 5. 🇲🇦 Marrakesh 6. 🇮🇸 Reykjavik 7. 🇮🇹 Palermo 8. 🇸🇪 Stockholm 9. 🇮🇱 Tel Aviv 10. 🇧🇷 Rio de Janeiro vs. 🏖️ Places the least attractive, lowest income & least educated visit 1. 🇹🇭 Pattaya 2. 🇻🇳 Da Nang 3. 🇱🇰 Colombo 4. 🇵🇭 Cebu 5. 🇮🇩 Jakarta 6. 🇲🇾 Kuala Lumpur 7. 🇵🇭 Manila 8. 🇦🇱 Tirana 9. 🇻🇳 Ho Chi Minh City 10. 🇹🇭 Phuket nomads.com/stats

This post is for software developers who feel betrayed by AI. And for everyone who is in the field of SW. "things changed, it’s not your weakness, it’s not that you are AI-pilled. It is just that our field is evolving in an incredible *and* painful (but also joyful) direction”.

@Jason I’ve gained 5 pounds watching your trip.

I'm a huge OnlyFlans fan

@appltrack The article says the M7 Ultra is coming in 2028 not 2029

Apple's M7 Ultra chip coming in 2029 is rumored to support 1.5TB of RAM. This would make the processor much more capable for on-device AI.

Great recommendation from @claybavor and reemphasizes @paulg and @shaneparrish’s thoughts on reading. The Wright Brothers were true entrepreneurs and the book can also be considered subtly about the power of thoughtful parenting. This may just be my personal lens. 🧐

@CinemaTweets1 :(

Sam Neill has passed away at 78. His passing comes suddenly after he stated in April that he was cancer free after fighting a rare form of blood cancer. I will miss this man deeply. He was an actor filled with heart and filled with joy. We lost a great one today. RIP to a true gentleman.

@justalexoki Still find it weird to think that some people don’t have it

how the hell do you write tweets if you don't have an internal monologue

If you could own any vehicle you wanted, what would it be?

@samhogan @OpenAI 😂

Thanks @OpenAI