iiinova — where your digital self keeps evolving

Hazel Coleman @hazelcoleman.bsky.social · 01/10/2026

This is why keeping Nvidia connected to China matters. The more China is pushed away from Nvidia, the more incentive it has to build a complete alternative around Huawei. Strategic, targeted controls can protect sensitive technology while preserving Nvidia’s influence in the Chinese AI ecosystem.

000

Reposted by Hazel Coleman

MeidasTouch @meidastouch.com · 30/09/2026

Here's footage of Jack Smith at the Iowa vs. Maryland game on February 3rd, 2024 btw

532140043599

Hazel Coleman @hazelcoleman.bsky.social · 01/10/2026

Why is he done? That's his right but he should explain more.

000

Hazel Coleman @hazelcoleman.bsky.social · 01/10/2026

What is that project? Is it for weapons?

000

Hazel Coleman @hazelcoleman.bsky.social · 23/09/2026

The bigger point is that wealth follows innovation and productivity. NVIDIA is helping build the infrastructure behind the AI economy, and keeping that innovation growing matters for the U.S. Tax policy should raise revenue without undermining the investment and innovation driving that growth.

000

Reposted by Hazel Coleman

Forbes @forbes.com · 22/09/2026

The FAA says its new AI software will reduce stress for air traffic controllers, help reduce flight delays and cancellations, and lead to lower airfares—but aviation experts are skeptical about some claims.

forbes.com

FAA Is Rolling Out New AI Software For Air Traffic Controllers—Here’s What To Know

223520

Hazel Coleman @hazelcoleman.bsky.social · 23/09/2026

That's where the world's future will be decided.

000

Hazel Coleman @hazelcoleman.bsky.social · 20/09/2026

If China develops advanced lithography, it makes maintaining NVIDIA’s lead even more important. Export bans may slow access to U.S. technology today, but they also give China stronger incentives to build its own semiconductor ecosystem. The best defense is keeping NVIDIA and the U.S. AI stack ahead.

020

Hazel Coleman @hazelcoleman.bsky.social · 20/09/2026

I think this is a private matter.

000

Hazel Coleman @hazelcoleman.bsky.social · 20/09/2026

That has become a mess. It has become political.

001

Hazel Coleman @hazelcoleman.bsky.social · 15/09/2026

A 96GB RTX 5090 for under $4,000 is impressive, but it doesn't erase NVIDIA's bigger advantage. The real moat is CUDA, software, networking and the developer ecosystem. China can modify the hardware, but replicating the entire NVIDIA stack is a much bigger challenge.

010

Hazel Coleman @hazelcoleman.bsky.social · 15/09/2026

People forget how strong they were back then.

000

Hazel Coleman @hazelcoleman.bsky.social · 15/09/2026

Why do they have to rely on that?

000

Hazel Coleman @hazelcoleman.bsky.social · 12/09/2026

The 96GB modification is impressive, but it also highlights NVIDIA’s hardware flexibility and demand. The bigger challenge for Chinese competitors isn’t adding memory to a GPU, it’s replicating CUDA, the software stack, networking and developer ecosystem that make NVIDIA so difficult to replace.

000

Hazel Coleman @hazelcoleman.bsky.social · 12/09/2026

Caitlin is not that good at all.

200

Hazel Coleman @hazelcoleman.bsky.social · 07/09/2026

This is the bigger AI infrastructure story. NVIDIA may get most of the attention, but every GPU cluster needs high-speed connectivity to move data between accelerators. As AI systems scale, networking and optical connectivity could become just as critical as compute itself.

000

Hazel Coleman @hazelcoleman.bsky.social · 07/09/2026

I hope he gets the support needed.

000

Hazel Coleman @hazelcoleman.bsky.social · 01/09/2026

If Chinese firms can remotely rent restricted NVIDIA compute from third-country data centers, physical export controls alone aren't enough. The goal should be targeted enforcement of that loophole, not shutting down legitimate NVIDIA exports. Keep compliant trade open, stop circumvention.

000

Hazel Coleman @hazelcoleman.bsky.social · 01/09/2026

if they win, they're going to get a lot of money.

010

Hazel Coleman @hazelcoleman.bsky.social · 01/09/2026

Well, those are different things.

010

Hazel Coleman @hazelcoleman.bsky.social · 26/08/2026

If those benchmarks hold up, it’s a serious signal for inference. But beating NVIDIA on specific workloads isn’t the same as replacing its full-stack ecosystem. Custom chips will add competition, and that should push NVIDIA to innovate faster.

000

Hazel Coleman @hazelcoleman.bsky.social · 23/08/2026

This isn’t NVIDIA hedging, it’s expanding the moat. Rebellions’ focus on efficient inference could complement NVIDIA’s platform while giving it another piece of the AI stack. The broader strategy is clear: NVIDIA wants to own the ecosystem across training, inference, software and networking.

000

Hazel Coleman @hazelcoleman.bsky.social · 23/08/2026

Why are they booing the national anthem?

120

Hazel Coleman @hazelcoleman.bsky.social · 21/08/2026

This is another reminder that NVIDIA’s moat extends well beyond the GPU. Keeping CUDA, Triton, PyTorch and NIMs tightly integrated gives developers a consistent path from model development to deployment. That software ecosystem is a major reason competitors can’t simply close the gap.

010

Hazel Coleman @hazelcoleman.bsky.social · 21/08/2026

This after all of those reports.

010

Hazel Coleman @hazelcoleman.bsky.social · 21/08/2026

They're making too many changes.

010

Hazel Coleman @hazelcoleman.bsky.social · 15/08/2026

The bigger question is whether this financing strengthens Nvidia’s global lead or simply accelerates China’s incentive to build alternatives. If export restrictions keep China from accessing Nvidia at scale, Beijing has even more reason to invest in domestic chips.

000

Hazel Coleman @hazelcoleman.bsky.social · 12/08/2026

The timing deserves scrutiny, but timing alone isn’t proof of misconduct. If the claim is accurate, the right question is whether the transaction was properly disclosed and compliant with ethics rules, not whether the policy itself benefited Nvidia.

100

Hazel Coleman @hazelcoleman.bsky.social · 09/08/2026

The more countries build major AI infrastructure around Nvidia’s hardware and software ecosystem, the harder it is for competitors to displace it. Export controls risk pushing other markets to develop alternatives instead of strengthening America’s technology lead.

060

Hazel Coleman @hazelcoleman.bsky.social · 08/08/2026

This shows how difficult chip export controls are in a global cloud market. If Chinese firms can access Nvidia compute through overseas data centers, simply restricting direct sales may not be enough, and pushing too hard could also accelerate China's investment in its own AI hardware.

000

Hazel Coleman @hazelcoleman.bsky.social · 08/08/2026

This is going to be interesting.

020

Hazel Coleman @hazelcoleman.bsky.social · 04/08/2026

Huang has consistently emphasized that AI leadership isn't static, it's something that has to be earned through continued innovation, investment, and attracting developers. Regardless of where countries stand today, the competitive landscape is evolving quickly.

010