Sentient Being
@schorndorfer.bsky.social
Reposted by Sentient Being
This was a remarkable book. Right at the intersection of my two interests these days. www.amazon.com/Chinese-Comp...
The Chinese Computer: A Global History of the Information Age
The Chinese Computer: A Global History of the Information Age - Kindle edition by Mullaney, Thomas S.. Download it once and read it on your Kindle device, PC, phones or tablets. Use features like bookmarks, note taking and highlighting while reading The Chinese Computer: A Global History of the Information Age.
www.amazon.com
December 26, 2024 at 4:16 PM
This was a remarkable book. Right at the intersection of my two interests these days. www.amazon.com/Chinese-Comp...
Reposted by Sentient Being
Yes, Americans are much richer than Japanese people
Yes, Americans are much richer than Japanese people
GDP is not a perfect guide to wealth, but it's pretty darn good.
www.noahpinion.blog
December 24, 2024 at 9:45 AM
Yes, Americans are much richer than Japanese people
Reposted by Sentient Being
Bluesky can be a fraught place to post about AI but it is worth noting that the buzz over o1 (& now o3) is not “hype.” We know o1 can actually do some very hard tasks (see my post) & o3 appears to represent a big further leap.
They aren’t AGI, but will matter. www.oneusefulthing.org/p/what-just-...
They aren’t AGI, but will matter. www.oneusefulthing.org/p/what-just-...
What just happened
A transformative month rewrites the capabilities of AI
www.oneusefulthing.org
December 22, 2024 at 10:05 AM
Bluesky can be a fraught place to post about AI but it is worth noting that the buzz over o1 (& now o3) is not “hype.” We know o1 can actually do some very hard tasks (see my post) & o3 appears to represent a big further leap.
They aren’t AGI, but will matter. www.oneusefulthing.org/p/what-just-...
They aren’t AGI, but will matter. www.oneusefulthing.org/p/what-just-...
Reposted by Sentient Being
a. yes training on the train set is legit.
b. but at the same time if you claim "general" intelligence then training for the task does make the claim much weaker.
c. why train on only 75% of the train set???
b. but at the same time if you claim "general" intelligence then training for the task does make the claim much weaker.
c. why train on only 75% of the train set???
December 22, 2024 at 4:00 PM
a. yes training on the train set is legit.
b. but at the same time if you claim "general" intelligence then training for the task does make the claim much weaker.
c. why train on only 75% of the train set???
b. but at the same time if you claim "general" intelligence then training for the task does make the claim much weaker.
c. why train on only 75% of the train set???
Reposted by Sentient Being
These kinds of headlines seem kind of dull and harmless now:
December 22, 2024 at 12:58 AM
These kinds of headlines seem kind of dull and harmless now:
Reposted by Sentient Being
What are the checks and balances on the power of Elon Musk?
What are the checks and balances on the power of Elon Musk?
The question we'll all be asking over the next year or more.
www.noahpinion.blog
December 21, 2024 at 9:05 AM
What are the checks and balances on the power of Elon Musk?
Reposted by Sentient Being
The polio vaccine works.
December 13, 2024 at 6:35 PM
The polio vaccine works.
Reposted by Sentient Being
Very glad to see this trend finally reversing
December 16, 2024 at 8:00 PM
Very glad to see this trend finally reversing
Reposted by Sentient Being
Mandating all-in upfront pricing is such an obviously good rule
December 18, 2024 at 1:56 PM
Mandating all-in upfront pricing is such an obviously good rule
Reposted by Sentient Being
I'll get straight to the point.
We trained 2 new models. Like BERT, but modern. ModernBERT.
Not some hypey GenAI thing, but a proper workhorse model, for retrieval, classification, etc. Real practical stuff.
It's much faster, more accurate, longer context, and more useful. 🧵
We trained 2 new models. Like BERT, but modern. ModernBERT.
Not some hypey GenAI thing, but a proper workhorse model, for retrieval, classification, etc. Real practical stuff.
It's much faster, more accurate, longer context, and more useful. 🧵
December 19, 2024 at 4:45 PM
I'll get straight to the point.
We trained 2 new models. Like BERT, but modern. ModernBERT.
Not some hypey GenAI thing, but a proper workhorse model, for retrieval, classification, etc. Real practical stuff.
It's much faster, more accurate, longer context, and more useful. 🧵
We trained 2 new models. Like BERT, but modern. ModernBERT.
Not some hypey GenAI thing, but a proper workhorse model, for retrieval, classification, etc. Real practical stuff.
It's much faster, more accurate, longer context, and more useful. 🧵
Reposted by Sentient Being
Honestly, Captain Highpants has no right to make fun of peoples looks anymore
December 18, 2024 at 6:03 PM
Honestly, Captain Highpants has no right to make fun of peoples looks anymore
Reposted by Sentient Being
while with LLMs, they are a fundamentally new thing, and thinking in terms of "classic" NLP shines the light on many cases where they are not-that-great (or not-that-great-yet), while distracting from considerations of their actual potential for radically new things.
December 16, 2024 at 11:49 PM
while with LLMs, they are a fundamentally new thing, and thinking in terms of "classic" NLP shines the light on many cases where they are not-that-great (or not-that-great-yet), while distracting from considerations of their actual potential for radically new things.
Reposted by Sentient Being
Just a reminder that none of the people who make LLMs, no matter how smart, actually know what specific tasks LLMs will be good or bad at. We are barely benchmarking these systems at all on any sorts of tasks.
You should explore in areas of your expertise to try to figure it out for your use cases.
You should explore in areas of your expertise to try to figure it out for your use cases.
December 16, 2024 at 3:14 AM
Just a reminder that none of the people who make LLMs, no matter how smart, actually know what specific tasks LLMs will be good or bad at. We are barely benchmarking these systems at all on any sorts of tasks.
You should explore in areas of your expertise to try to figure it out for your use cases.
You should explore in areas of your expertise to try to figure it out for your use cases.
Reposted by Sentient Being
At least five interesting things: Will of the Masses edition (#55)
At least five interesting things: Will of the Masses edition (#55)
Abandoning X; Drone world; Asian voters' rightward shift; China and the Global South; Polycrisis; Illegal immigration's fiscal impact; Iranian power
www.noahpinion.blog
December 16, 2024 at 6:51 AM
At least five interesting things: Will of the Masses edition (#55)
Reposted by Sentient Being
Why isn't Functional Programming the Norm?
www.youtube.com/watch?v=QyJ...
www.youtube.com/watch?v=QyJ...
December 12, 2024 at 6:00 PM
Why isn't Functional Programming the Norm?
www.youtube.com/watch?v=QyJ...
www.youtube.com/watch?v=QyJ...
Reposted by Sentient Being
Iraqi architecture, Karbala 🇮🇶
December 12, 2024 at 7:08 AM
Iraqi architecture, Karbala 🇮🇶
Reposted by Sentient Being
what a splendid example of natural and useful LLM/UX integration!
The latest version of Obsidian Web Clipper adds Interpreter which lets you extract page content and manipulate it using natural language.
This is super useful to create consistent data structures for research within your vault. You can run it locally using Ollama.
help.obsidian.md/web-clipper/...
This is super useful to create consistent data structures for research within your vault. You can run it locally using Ollama.
help.obsidian.md/web-clipper/...
December 12, 2024 at 9:39 AM
what a splendid example of natural and useful LLM/UX integration!
Reposted by Sentient Being
The latest version of Obsidian Web Clipper adds Interpreter which lets you extract page content and manipulate it using natural language.
This is super useful to create consistent data structures for research within your vault. You can run it locally using Ollama.
help.obsidian.md/web-clipper/...
This is super useful to create consistent data structures for research within your vault. You can run it locally using Ollama.
help.obsidian.md/web-clipper/...
December 10, 2024 at 9:35 PM
The latest version of Obsidian Web Clipper adds Interpreter which lets you extract page content and manipulate it using natural language.
This is super useful to create consistent data structures for research within your vault. You can run it locally using Ollama.
help.obsidian.md/web-clipper/...
This is super useful to create consistent data structures for research within your vault. You can run it locally using Ollama.
help.obsidian.md/web-clipper/...
Reposted by Sentient Being
I genuinely think if the companies were named "Heavy Metals Amalgamated" and "BTEK Industries" instead of US/Nippon Steel this deal could have survived
okay, lol
*BIDEN SET TO FINALLY KILL NIPPON STEEL PURCHASE OF US STEEL
*BIDEN SET TO FINALLY KILL NIPPON STEEL PURCHASE OF US STEEL
This sort of stuff is evidence IMO that he's going to get ~nothing meaningful done.
December 11, 2024 at 6:23 PM
I genuinely think if the companies were named "Heavy Metals Amalgamated" and "BTEK Industries" instead of US/Nippon Steel this deal could have survived
Reposted by Sentient Being
The home includes multiple windows and a sliding glass door to welcome in the natural light. As with all of our models, everything can be customized to suite your specific needs.
frontiertinyhomes.com/perpetua/
frontiertinyhomes.com/perpetua/
Perpetua - Frontier Tiny Homes
The Perpetua is a two-story tiny home with up to four bedrooms. The main level features a master bedroom with a closet, a full bathroom with a walk-in shower,
frontiertinyhomes.com
December 11, 2024 at 11:02 AM
The home includes multiple windows and a sliding glass door to welcome in the natural light. As with all of our models, everything can be customized to suite your specific needs.
frontiertinyhomes.com/perpetua/
frontiertinyhomes.com/perpetua/
Reposted by Sentient Being
The new ChatGPT Canvas mode adds a 2nd way to execute Python inside ChatGPT - using Pyodide+WASM, where Code Interpreter uses server-side Python in a Kubernetes sandbox
And Pyodide can make network requests!
I wrote about how exciting and confusing this is here: simonwillison.net/2024/Dec/10/...
And Pyodide can make network requests!
I wrote about how exciting and confusing this is here: simonwillison.net/2024/Dec/10/...
ChatGPT Canvas can make API requests now, but it’s complicated
Today’s 12 Days of OpenAI release concerned ChatGPT Canvas, a new ChatGPT feature that enables ChatGPT to pop open a side panel with a shared editor in it where you …
simonwillison.net
December 10, 2024 at 9:53 PM
The new ChatGPT Canvas mode adds a 2nd way to execute Python inside ChatGPT - using Pyodide+WASM, where Code Interpreter uses server-side Python in a Kubernetes sandbox
And Pyodide can make network requests!
I wrote about how exciting and confusing this is here: simonwillison.net/2024/Dec/10/...
And Pyodide can make network requests!
I wrote about how exciting and confusing this is here: simonwillison.net/2024/Dec/10/...
Reposted by Sentient Being
Insurance companies aren't the main villain of the U.S. health system
Insurance companies aren't the main villain of the U.S. health system
It's mostly the providers overcharging you, not the middlemen.
www.noahpinion.blog
December 9, 2024 at 9:15 PM
Insurance companies aren't the main villain of the U.S. health system
Reposted by Sentient Being
Reposted by Sentient Being
I can't begin to describe how life-changing this new project, ShellSage, has been for me over the last few weeks.
ShellSage is an LLM that lives in your terminal. It can see what directory you're in, what commands you've typed, what output you got, & your previous AI Q&A's.🧵
ShellSage is an LLM that lives in your terminal. It can see what directory you're in, what commands you've typed, what output you got, & your previous AI Q&A's.🧵
As R&D staff @ answer.ai, I work a lot on boosting productivity with AI. A common theme that always comes up is the combination of human+AI. This combination proved to be powerful in our new project ShellSage, which is an AI terminal buddy that learns and teaches with you. A 🧵
December 5, 2024 at 8:30 PM
I can't begin to describe how life-changing this new project, ShellSage, has been for me over the last few weeks.
ShellSage is an LLM that lives in your terminal. It can see what directory you're in, what commands you've typed, what output you got, & your previous AI Q&A's.🧵
ShellSage is an LLM that lives in your terminal. It can see what directory you're in, what commands you've typed, what output you got, & your previous AI Q&A's.🧵
Reposted by Sentient Being
Portland Airport’s new main terminal features an innovative mass timber roof canopy designed by ZGF Architects. Spanning nine acres, the expansive structure uses locally sourced Douglas Fir glulam panels, creating a sustainable and regionally authentic design.
Photo: Ema Peter & Dror Baldinger FAI
Photo: Ema Peter & Dror Baldinger FAI
December 5, 2024 at 6:38 PM
Portland Airport’s new main terminal features an innovative mass timber roof canopy designed by ZGF Architects. Spanning nine acres, the expansive structure uses locally sourced Douglas Fir glulam panels, creating a sustainable and regionally authentic design.
Photo: Ema Peter & Dror Baldinger FAI
Photo: Ema Peter & Dror Baldinger FAI