
Literal rock star developer. JavaScript, music, Japanese, and serial commas. 日本語でもOKです。Author and maintainer of the EJS templating lib. Developer at @usebland.
San Francisco, California USA
Joined February 2007
- Tweets22.5K
- Following235
- Followers2.3K
- Likes10.3K
holy shit, they literally created an entire world in minecraft.
i recreated THE ENTIRE WORLD in minecraft at 1:1 scale. every building on earth, including your house!
to do this, i trained a SOTA world-to-minecraft generator on maps data, complete with furnished home interiors!
play and check out your neighborhood: worldcraft.gg
My god, the comments, so brutal. I don't know why these fucking idiots even post on Twitter anymore.
Texas is now a blue state.
But Texas did not fundamentally change who we are. We did not wake up on election day and suddenly become a tofu-eating, Birkenstock-wearing, vegan, gun-controlling, commie, lefty, pink state. We are not California.
People were frustrated with the price of gas and food, and that frustration is real. As Republicans, we need to listen.
Here’s the good news: we are going to turn Texas around. We are going to save Texas.
25
Introducing Jev @typesafeai powered jobs-api
Try it at treg.to/jobs-api
- One API to any jobs data
- $0.03 to run a full job hunt
- Real salary & Hiring manager data
- Plugin to any agent for your dream job
Fully open source, 0% markup
Git Repo below 👇
All that is old is new again: web.archive.org/web/20220823…
many such cases
ORMs are great, but at scale you need to pay attention to the queries they generate.
very easy to end up with inefficient queries
72
Matthew Eernisse retweeted
Relaxing with a little bowl of cereal at 11pm in a safe home while scrolling the worldcontent is a luxury beyond your ancestors wildest dreams
Introducing Tofu-1, Bland’s new speech-to-speech model.
Every sound you hear in this video, the caller, the agent, and the background sounds, was generated by Tofu-1.
High-stakes calls don’t happen in quiet rooms. They happen in storms, on the side of highways, with kids crying in the background. Our model was trained on the callers that enterprises talk to every day, so your agents can handle them too.
Join the research preview: bland.ai/speech-to-speech?ut…
"Stop replying to all."
kanmail.io/blog/reply-all-st…
(This happened to me at Microsoft.)
1
3
175
Want to hear what the future of AI voice sounds like?
Request Early Access: bland.ai/speech-to-speech#re…
Today we’re introducing Tofu-1, Bland’s speech-to-speech model.
The highest-stakes calls rarely happen in good conditions: from moving cars and hospital rooms, over weak signal, with several people talking at once.
Tofu-1 is built for those calls. It listens while it speaks, reasons while it talks, and can tell a caller apart from everything else on a line, like a TV or a crying baby.
Join the preview: bland.ai/speech-to-speech?ut…
This guy's pretty good.
This is a message for @realDonaldTrump.
Donald, this is on the table. Guarantee healthcare for every American, from cradle to grave.
2
94
Matthew Eernisse retweeted
This post by Anthropic is massively overlooked, and these are my biggest takeaways:
- Claude now leads 26% of Anthropic's AI R&D work. In February it was under 1%
- AI collaborates on or leads over 90% of that R&D work. In August 2025, going by the chart, it was about 4%
- About 30,000 AI agents do research and engineering at Anthropic at any one time (on one platform)
- Those agents made over a billion decisions in one month. 1 in 47,000 got blocked
- Monitors flag ~100,000 agent transcripts a week, and only ~50 reach a human
- To build the index, Claude agents read staff Slack and docs and mapped 15,000 R&D tasks
- Claude agreed with humans more often than humans agreed with each other (59% vs 35%)
- Agents get their own identities that survive model upgrades. They coordinate on an open message board, so they can check each other's work
anthropic.com/institute/meas…
I keep saying this: seems high. reuters.com/world/us/trump-a…
1
2
69

















































