@ghchinoy.bsky.social
"a statistics engine that creates sentences that sound accurate"
Reposted
I haven’t been able to stop thinking about this NYT article about programming work at Amazon (https://www.nytimes.com/2025/05/25/business/amazon-ai-coders.html) because it perfectly illustrates how “productivity” trends in AI in the workplace benefit so few.
June 3, 2025 at 3:24 PM
I haven’t been able to stop thinking about this NYT article about programming work at Amazon (https://www.nytimes.com/2025/05/25/business/amazon-ai-coders.html) because it perfectly illustrates how “productivity” trends in AI in the workplace benefit so few.
Catastrophic forgetting weekend - willingly closed all my chrome tabs after setting to not reopen. Awaiting the wonders of Monday with eager eyes.
May 18, 2025 at 10:32 PM
Catastrophic forgetting weekend - willingly closed all my chrome tabs after setting to not reopen. Awaiting the wonders of Monday with eager eyes.
Reposted
@pydantic.dev's FastA2A! (A2A is Google's Agent2Agent Protocol, which complements MCP (developers.googleblog.com/en/a2a-a-new...).
ai.pydantic.dev/a2a/
ai.pydantic.dev/a2a/
May 13, 2025 at 10:23 PM
@pydantic.dev's FastA2A! (A2A is Google's Agent2Agent Protocol, which complements MCP (developers.googleblog.com/en/a2a-a-new...).
ai.pydantic.dev/a2a/
ai.pydantic.dev/a2a/
Based on all the hype in my news feed, I figured Murderbot was over. It hasn't even debuted and people have season one reviews. I guess I'll just reread the books.
May 13, 2025 at 12:18 PM
Based on all the hype in my news feed, I figured Murderbot was over. It hasn't even debuted and people have season one reviews. I guess I'll just reread the books.
Reposted
Turn any gRPC api into an #mcp server: github.com/redpanda-dat... ... watch a live demo here: youtu.be/gIFN8NlV8vk?... ... let's go! ... apache 2 licensed
April 29, 2025 at 10:03 PM
Turn any gRPC api into an #mcp server: github.com/redpanda-dat... ... watch a live demo here: youtu.be/gIFN8NlV8vk?... ... let's go! ... apache 2 licensed
Reposted
A mob of Hasidic men in Brooklyn chase down a woman who came to protest Ben-Gvir.
This sort of reaction has become normalized in NYC and LA, where violence against protestors has spiked enormously since October 7th.
This largely isn't covered or addressed by media, despite how widespread it is.
This sort of reaction has become normalized in NYC and LA, where violence against protestors has spiked enormously since October 7th.
This largely isn't covered or addressed by media, despite how widespread it is.
April 25, 2025 at 8:57 PM
A mob of Hasidic men in Brooklyn chase down a woman who came to protest Ben-Gvir.
This sort of reaction has become normalized in NYC and LA, where violence against protestors has spiked enormously since October 7th.
This largely isn't covered or addressed by media, despite how widespread it is.
This sort of reaction has become normalized in NYC and LA, where violence against protestors has spiked enormously since October 7th.
This largely isn't covered or addressed by media, despite how widespread it is.
It's just the trailer but an oddly tone deaf attempt at dark humor. Would rather see a Careless People movie. Or read it again.
search.app/aLg3Z
search.app/aLg3Z
Mountainhead | Official Teaser | Max
Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.
search.app
April 24, 2025 at 1:12 AM
It's just the trailer but an oddly tone deaf attempt at dark humor. Would rather see a Careless People movie. Or read it again.
search.app/aLg3Z
search.app/aLg3Z
Reposted
This is a great thread, and it’s why every time someone says, cynically “well, do you think the courts are going to save us?” I think, “not on their own, no, but we aren’t going to get saved without a real legal apparatus.”
alright, i'm asking in seriousness. do folks know what lynching is? do folks know the *legal* issues with lynching? it's actually more complicated than you think, and i'm not a legal scholar so i don't want to step out of my lane--BUT I'VE READ BOOKS.
April 19, 2025 at 6:23 PM
This is a great thread, and it’s why every time someone says, cynically “well, do you think the courts are going to save us?” I think, “not on their own, no, but we aren’t going to get saved without a real legal apparatus.”
Reposted
alright, i'm asking in seriousness. do folks know what lynching is? do folks know the *legal* issues with lynching? it's actually more complicated than you think, and i'm not a legal scholar so i don't want to step out of my lane--BUT I'VE READ BOOKS.
April 18, 2025 at 8:49 PM
alright, i'm asking in seriousness. do folks know what lynching is? do folks know the *legal* issues with lynching? it's actually more complicated than you think, and i'm not a legal scholar so i don't want to step out of my lane--BUT I'VE READ BOOKS.
I don't know enough about linguistics to know if there are languages that have more emphasis on time duration (aspect) than ontera or if there's an analogy for the behavior described here (the "green banapple" for example). It'd be super interesting to know.
arxiv.org/abs/2504.03933
arxiv.org/abs/2504.03933
Language Models Are Implicitly Continuous
Language is typically modelled with discrete sequences. However, the most successful approaches to language modelling, namely neural networks, are continuous and smooth function approximators. In this...
arxiv.org
April 20, 2025 at 2:46 PM
I don't know enough about linguistics to know if there are languages that have more emphasis on time duration (aspect) than ontera or if there's an analogy for the behavior described here (the "green banapple" for example). It'd be super interesting to know.
arxiv.org/abs/2504.03933
arxiv.org/abs/2504.03933
Fantastic post on how the portrayal of Celebrimbor in the TV shows misses the moral points of the text
From @bretdevereaux.bsky.social
acoup.blog/2025/04/18/c...
From @bretdevereaux.bsky.social
acoup.blog/2025/04/18/c...
Collections: Why Celebrimbor Fell but Boromir Conquered: the Moral Universe of Tolkien
This week (and probably next) I want to talk a bit more Tolkien, but in a somewhat different vein from normal. Rather than discussing the historicity of Tolkien’s world or adaptations of it, …
acoup.blog
April 19, 2025 at 2:21 PM
Fantastic post on how the portrayal of Celebrimbor in the TV shows misses the moral points of the text
From @bretdevereaux.bsky.social
acoup.blog/2025/04/18/c...
From @bretdevereaux.bsky.social
acoup.blog/2025/04/18/c...
Reposted
Save the date! I'm moderating an @awmuseum.bsky.social discussion about creating faith systems in fictional worlds with @nkjemisin.bsky.social @nnedi.bsky.social @nghivo.bsky.social & Matthew Kirby on 4/22 at 5:15pm CST at the Chicago Public Library. americanwritersmuseum.org/program-cale...
April 3, 2025 at 5:52 PM
Save the date! I'm moderating an @awmuseum.bsky.social discussion about creating faith systems in fictional worlds with @nkjemisin.bsky.social @nnedi.bsky.social @nghivo.bsky.social & Matthew Kirby on 4/22 at 5:15pm CST at the Chicago Public Library. americanwritersmuseum.org/program-cale...
SC > SAR > SIAR > ASI
Missing "superhuman mansplainer"
Missing "superhuman mansplainer"
April 3, 2025 at 11:59 PM
SC > SAR > SIAR > ASI
Missing "superhuman mansplainer"
Missing "superhuman mansplainer"
This is odd to me that they're all so similar. There has to be at least one macroeconomics book in the common crawl.
guess where they got their weird trade deficit math from?
i went to the pit for y'all and brought back the screenshots with alt text
i went to the pit for y'all and brought back the screenshots with alt text
![chatgpt log:
What would be an easy way to calculate the tariffs that should be imposed on other countries so that the US is on even-playing fields when it comes to trade deficit? Set minimum at 10%.
Reasoned about trade tariffs and deficits for 10 seconds v
A straightforward (if naïve) method is to set the tariff rate for each trading partner equal to the percentage share of the trade deficit relative to that country's total imports-always no less than 10%. In formula form:
Tariff Rate (%) = max(10, (Trade Deficit ÷ Total Imports) *
100)
For example, if the US runs a $20 billion deficit on $100 billion worth of imports from a country, the calculated rate is (20/100)×100 = 20%. If the deficit were only 5%, you'd still impose a 10% floor.
This method ignores the intricate dynamics of international trade-such as elasticities, retaliatory measures, and supply chain nuances-but it provides a blunt, proportional rule to "level the playing field." One might implement it in Python like so:
[code]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:iyz5zf463ic52vqbonyu2ebu/bafkreigbalqxjorllnpelbwe5kcla5fdedxvgv3cicxdsrbdxarodif5ba@jpeg)
![Grok:
the deficit. The deficit as a percentage of total trade is
$295B/$582B = 50.69%. Using the formula Tariff% = 10% +
(Deficit% × Scaling Factor), we choose a scaling factor of 50 for illustration:
• Tariff = 10% + (50.69% × 50) = 10% + 25.345% = 35.345%, rounded to 35.3%.
This means a 35.3% tariff on Chinese goods could be proposed to offset the imbalance, aiming to reduce imports and encourage domestic production or exports.
Extending to Other Countries
To provide a broader view, here's a table of top US trading partners with 2024 deficits, using data from The Motley Fool and similar calculations (assuming scaling factor of 50 for consistency):
[data table headings]
Country
Deficit
(USD
Billion)
Total
Trade
(USD
Billion)
Deficit %
Calculate d Tariff %
China
-295
582
50.69%
35.3%
Mexico
-172
840
20.48%
20.2%
Vietnam
-123
(Data incomplet e, estimate)
~30%
25.0%
Ireland
-87
(Data incomplet e, estimate)
~25%
22.5%
Germany
-85
(Data incomplete estimate)
~20%
20.0%](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:iyz5zf463ic52vqbonyu2ebu/bafkreietqlxsi5akdrvtp2rwqpc6kt6g6bbb2dppqcfn2j4q7xkofn4f4y@jpeg)
April 3, 2025 at 5:06 AM
This is odd to me that they're all so similar. There has to be at least one macroeconomics book in the common crawl.
Reposted
The second (non-joke) blog post is publishing now, with details:
tailscale.com/blog/plan9-p...
We'll be in Google Meet in ~45m here to chat: ftp.plan9.ts.net/webinar
tailscale.com/blog/plan9-p...
We'll be in Google Meet in ~45m here to chat: ftp.plan9.ts.net/webinar
Porting Tailscale to Plan 9
Securely connect to anything on the internet with Tailscale. Built on WireGuard®️, Tailscale enables you to make finely configurable connections, secured end-to-end according to zero trust principles,...
tailscale.com
April 2, 2025 at 3:18 PM
The second (non-joke) blog post is publishing now, with details:
tailscale.com/blog/plan9-p...
We'll be in Google Meet in ~45m here to chat: ftp.plan9.ts.net/webinar
tailscale.com/blog/plan9-p...
We'll be in Google Meet in ~45m here to chat: ftp.plan9.ts.net/webinar
Reposted
Reposted
Reposted
Join us for a #MAIHT3k livestream where @mmitchell.bsky.social will breakdown the latest "agentic AI" nonsense. See also the preprint she coauthored on the topic.
⏰ Tomorrow (Monday March 31st) at 12pm pacific
➡️ We'll see you at www.twitch.tv/dair_institute
📑 arxiv.org/abs/2502.02649
⏰ Tomorrow (Monday March 31st) at 12pm pacific
➡️ We'll see you at www.twitch.tv/dair_institute
📑 arxiv.org/abs/2502.02649
March 31, 2025 at 12:27 AM
Join us for a #MAIHT3k livestream where @mmitchell.bsky.social will breakdown the latest "agentic AI" nonsense. See also the preprint she coauthored on the topic.
⏰ Tomorrow (Monday March 31st) at 12pm pacific
➡️ We'll see you at www.twitch.tv/dair_institute
📑 arxiv.org/abs/2502.02649
⏰ Tomorrow (Monday March 31st) at 12pm pacific
➡️ We'll see you at www.twitch.tv/dair_institute
📑 arxiv.org/abs/2502.02649
Reposted
Whenever @hamel.bsky.social drops a new banger of a blog post I make sure to forward it to the rest of my team 🚀 And I encourage everyone working with productionalizing LLM systems to do the same.
hamel.dev/blog/posts/f...
hamel.dev/blog/posts/f...
A Field Guide to Rapidly Improving AI Products – Hamel’s Blog
Evaluation methods, data-driven improvement, and experimentation techniques from 30+ production implementations.
hamel.dev
March 26, 2025 at 1:50 PM
Whenever @hamel.bsky.social drops a new banger of a blog post I make sure to forward it to the rest of my team 🚀 And I encourage everyone working with productionalizing LLM systems to do the same.
hamel.dev/blog/posts/f...
hamel.dev/blog/posts/f...
Reposted
It's a fun sci-fi trope: learn an alien language and acquire superpowers, just like if you'd been zapped by a cosmic ray. But what's the linguistics behind this idea?
This month we get enthusiastic about the science and fiction of linguistic relativity!
This month we get enthusiastic about the science and fiction of linguistic relativity!
102: The science and fiction of Sapir-Whorf
It's a fun science fiction trope: learn a mysterious alien language and acquire superpowers, just like if you'd been zapped by a cosmic ray or bitten by a radioactive spider. But what's the linguistic
soundcloud.com
March 21, 2025 at 1:46 AM
It's a fun sci-fi trope: learn an alien language and acquire superpowers, just like if you'd been zapped by a cosmic ray. But what's the linguistics behind this idea?
This month we get enthusiastic about the science and fiction of linguistic relativity!
This month we get enthusiastic about the science and fiction of linguistic relativity!
Reposted
Yet another brilliant piece by danah boyd
Policymakers and technologists alike are stuck in a “solutionism” trap—but particularly with AI, we need to focus on intervention, not “solutions,” argues danah boyd. An interventionist framework ensures a more iterative and non-deterministic approach to shaping AI futures.
We Need an Interventionist Mindset | TechPolicy.Press
To stabilize and enhance democratic practices, danah boyd advocates for an interventionist approach to AI regulation.
buff.ly
March 28, 2025 at 9:31 PM
Yet another brilliant piece by danah boyd