its more fun to compute
August 7, 2025 at 5:52 PM
its more fun to compute
its more fun to compute
February 5, 2025 at 9:10 PM
its more fun to compute
"orbital compute"
Right now, launch costs drive up orbital compute to roughly four times the price of terrestrial, SemiAnalysis wrote. But the cost difference could fall to 30% within five years.
June 12, 2026 at 8:59 PM
"orbital compute"
godot 4.8 exposed mesh buffers to gpu compute. all sorts of cool procedural shit is possible now
September 21, 2026 at 11:13 PM
godot 4.8 exposed mesh buffers to gpu compute. all sorts of cool procedural shit is possible now
It's a meet-compute
og terminator is basically a romcom
I watched Terminator for the first time tonight, and while it’s a solid movie, I do think it’s funny that the premise is basically “Hey, how’d you like to go back in time and bang my mom”
February 18, 2024 at 3:53 AM
It's a meet-compute
Compute is not a noun.
January 19, 2026 at 5:42 AM
Compute is not a noun.
you can genuinely watch the news anchor not compute in real time
March 5, 2026 at 11:05 PM
you can genuinely watch the news anchor not compute in real time
this is going to burn a shitton of compute
SCOOP: OpenAI is preparing to launch a stand-alone app for its video generation AI model Sora 2, WIRED has learned.
Now what exactly will that look like? Here's what we know.
From @zoeschiffer.bsky.social and @lmatsakis.bsky.social:
Now what exactly will that look like? Here's what we know.
From @zoeschiffer.bsky.social and @lmatsakis.bsky.social:
OpenAI Is Preparing to Launch a Social App for AI-Generated Videos
The platform appears to closely resemble TikTok and is powered by Sora 2, OpenAI's latest video generation model.
www.wired.com
September 29, 2025 at 8:23 PM
this is going to burn a shitton of compute
How high are OpenAI’s compute costs? Possibly a lot higher than we thought on.ft.com/43qoQ2L
How high are OpenAI’s compute costs? Possibly a lot higher than we thought
Inference inferred, revenue reconstructed, cash burn quantified
on.ft.com
November 12, 2025 at 5:46 PM
How high are OpenAI’s compute costs? Possibly a lot higher than we thought on.ft.com/43qoQ2L
Trans girls using all the compute
[0008890] Lock 4 puppys in a room and the FPS goes down substantially
February 18, 2026 at 2:42 PM
Trans girls using all the compute
Wait, the shoe company...?
*ALLBIRDS ANNOUNCES EXPANSION INTO AI COMPUTE INFRASTRUCTURE
*ALLBIRDS ANNOUNCES EXPANSION INTO AI COMPUTE INFRASTRUCTURE
April 15, 2026 at 12:01 PM
Wait, the shoe company...?
*ALLBIRDS ANNOUNCES EXPANSION INTO AI COMPUTE INFRASTRUCTURE
*ALLBIRDS ANNOUNCES EXPANSION INTO AI COMPUTE INFRASTRUCTURE
Cost spikes without new resources.
Azure billing jumped 8% — likely from short-lived high-cost compute used for data extraction, then deleted.
Azure billing jumped 8% — likely from short-lived high-cost compute used for data extraction, then deleted.
April 18, 2025 at 12:12 AM
Cost spikes without new resources.
Azure billing jumped 8% — likely from short-lived high-cost compute used for data extraction, then deleted.
Azure billing jumped 8% — likely from short-lived high-cost compute used for data extraction, then deleted.
we need some sort of Statcast for saliva to compute what Dak was aiming for
September 5, 2025 at 5:27 PM
we need some sort of Statcast for saliva to compute what Dak was aiming for
Yes. Anthropic has made $5bn in its entire existence up until March. It also spent $10bn on compute. It's raised $60bn (with $30bn of that arriving Feb 12), which suggests other costs - or prepayments/deposits for compute are brutal
So less than GTA V or Genshin Impact? The technology that consumes all the capital, water, and energy in the world is being outperformed by gacha games?
March 10, 2026 at 10:34 PM
Yes. Anthropic has made $5bn in its entire existence up until March. It also spent $10bn on compute. It's raised $60bn (with $30bn of that arriving Feb 12), which suggests other costs - or prepayments/deposits for compute are brutal
if they can fit "macbook pro levels of compute" in an iphone why don't they just fill a macbook pro with that compute density
September 9, 2025 at 5:49 PM
if they can fit "macbook pro levels of compute" in an iphone why don't they just fill a macbook pro with that compute density
She yotta on my flop til i compute
January 7, 2026 at 5:07 PM
She yotta on my flop til i compute
Correct. “Peaceful transfer” does not compute. It’s a stupid thing to even say.
You cannot have a “peaceful transfer” to state-sponsored violence.
November 21, 2024 at 1:26 AM
Correct. “Peaceful transfer” does not compute. It’s a stupid thing to even say.
Allbirds, the AI compute infrastructure solutions provider?
April 15, 2026 at 12:15 PM
Allbirds, the AI compute infrastructure solutions provider?
finally a source for all the frankly bizarre claims around LLM water "use"
literally nothing about LLM-related compute uses more water than any other compute activity, focus on Actually Objectionable bits around LLMs like degraded infosphere
literally nothing about LLM-related compute uses more water than any other compute activity, focus on Actually Objectionable bits around LLMs like degraded infosphere
some researchers at uc riverside wrote a paper estimating water use, including indirect use from power generation. they counted hydroelectic water use as "used" water. their worst case assumes Washington State where power is mostly hydroelectric, other pessimistic things. they gave quotes to WAPO
May 9, 2025 at 10:39 PM
finally a source for all the frankly bizarre claims around LLM water "use"
literally nothing about LLM-related compute uses more water than any other compute activity, focus on Actually Objectionable bits around LLMs like degraded infosphere
literally nothing about LLM-related compute uses more water than any other compute activity, focus on Actually Objectionable bits around LLMs like degraded infosphere
spending $45,000 in compute to generate a version of this where we lose
First trailer for Darren Aranofsky's new AI animated series 'On This Day... 1776'
• Tells short narrative stories about the Revolutionary War
• Uses Gen AI tools, including tech made by Google DeepMind
• Has SAG voice actors
• Tells short narrative stories about the Revolutionary War
• Uses Gen AI tools, including tech made by Google DeepMind
• Has SAG voice actors
January 30, 2026 at 6:13 AM
spending $45,000 in compute to generate a version of this where we lose
"OK, so these models need ridiculous amounts of compute"
"Sounds expensive"
"Yes, we're spending trillions to build data centers"
"Cool, cool, what are the plans to make a return there?"
"TBD, oh and by the way we're working on making these models more compute efficient"
"So, fewer DCs?"
"No."
"Sounds expensive"
"Yes, we're spending trillions to build data centers"
"Cool, cool, what are the plans to make a return there?"
"TBD, oh and by the way we're working on making these models more compute efficient"
"So, fewer DCs?"
"No."
September 11, 2026 at 12:28 AM
"OK, so these models need ridiculous amounts of compute"
"Sounds expensive"
"Yes, we're spending trillions to build data centers"
"Cool, cool, what are the plans to make a return there?"
"TBD, oh and by the way we're working on making these models more compute efficient"
"So, fewer DCs?"
"No."
"Sounds expensive"
"Yes, we're spending trillions to build data centers"
"Cool, cool, what are the plans to make a return there?"
"TBD, oh and by the way we're working on making these models more compute efficient"
"So, fewer DCs?"
"No."
Can a deer even compute?
@lafawn.com
@lafawn.com
September 21, 2025 at 11:19 AM
Can a deer even compute?
@lafawn.com
@lafawn.com
