TC

[TC @TeaCrowl]

@SunWeatherMan @sunfellow How long until pole flip ballpark?

[Tesla Hype @TeslaHype]

Elon on TeraFab: "I have no idea how to make chips… but we'll figure it out." Same as cars and rockets.

[Elon Musk @elonmusk]

Electricity production is the best metric for the true strength of any large-scale economy imo

[Zephyr @zephyr_z9]

It's pretty clear that Nvidia will compete head on with the frontier labs now
Jensen is pissed because labs are developing their own ASICs and undercutting Nvidia

[Negligible Capital @negligible_cap]

*NVIDIA IN TALKS TO ACQUIRE REFLECTION AI: FT

What’re the odds reflection can reach the frontier with near bottomless compute? Net negative for OpenAI and Anthropic in the long run I’d imagine

Jensen’s also already invested $800m into Reflection

[Sean Florez @seanf1orez]

why is this cool af?

Cool idea: put a living area inside, and autonomously go from city to city.

[TC @TeaCrowl]

@trinity_314 How do you know this. Genuinely

[TC @TeaCrowl]

@u1tra_instinct So wait is the 64gb Mac Studio like speeding up the Spark?

[TC @TeaCrowl]

@ashxhart Sorry I’m a total novice. Does this mean you could in theory train a local model to become smarter than it already is?

I’m using Qwen 3.8 flash next with Tensorfold on a single spark it’s amazing

And would you use it still with something like Hermes?

[TC @TeaCrowl]

@Rimusz @MiaAI_lab @bot @Tailscale Isn’t this a major security risk?

[nicksaraev @nicksaraev]

I am not joking when I say this.

Traditional game development is over.

I asked Opus-5.5 to make Attack on Titan in Minecraft. Took ~1hr and cost less than $25 start to finish.

The end result is extremely fun. And multiplayer!

Guide + link to play below.

[TC @TeaCrowl]

@Alibaba_Qwen WHEN QWEN 4

[TC @TeaCrowl]

@Alibaba_Qwen Any update on Qwen 4? :)

[TC @TeaCrowl]

@bscholl I use this on the 757 at work.

Miserable. Turbulence plus a route change is honestly a threat.

[TC @TeaCrowl]

@ashxhart It’s like almost there.

Like 90% what I’m looking for in a AI model

I wish I had more sparks lol

I’m still having to to use frontier here and there. I loveeee Tensorfold

[TC @TeaCrowl]

@aj_geo_analysis You need a Key to explain the map

[Jared @RevJaredJones]

America is awesome because an F18 pilot can eject, walk down the road on an iphone, and a local can stand right next to him talking as if he knows exactly whats going on and says, "Damn bud, you hit the tree, you need to fly higher."

[TC @TeaCrowl]

@elonmusk @DiamondEyesFox That’s only possible if humanity has more people. Our biggest problem is amount of humans needed

[Mia @MiaAI_lab]

Time for some love for single DGX Spark owners.

Qwen3.8 Flash Next on TensorFold v1.0 is out ✨

- Runs on TensorFold Zig.
- Swapped the MLX quant for Int4-AutoRound.
- Much more reliable than before.
- Squashed many bugs.

Performance:
- 64.4 tok/s on prose, single stream
- 140 tok/s on prose, 4 concurrent streams
- 200 tok/s on prose, 8 concurrent streams
- Prefill ~2500-2600 tok/s.

Still the best model you can run on a single DGX Spark. Nothing else comes close. @ashxhart

Get it here:
https://t.co/DZy94NF7zm

[TC @TeaCrowl]

Mercy to the guilty is cruelty to the innocent

[Complex @Complex]

A father threw a chair toward a judge after the driver who killed his 2 year old daughter and both her grandparents was sentenced to just 120 hours of community service.

[H/T: insidehistory]