ryanbehdad.bsky.social
@ryanbehdad.bsky.social
Reposted by ryanbehdad.bsky.social
You can just take academic papers and paste them into Gemini 2.5/ChatGPT o3/Claude 4 with the prompt "build me a game based on this paper, make it interesting and thematic but still conveying key findings" and get a tiny working educational game. (In this case, I used Gemini)
May 23, 2025 at 4:37 PM
You can just take academic papers and paste them into Gemini 2.5/ChatGPT o3/Claude 4 with the prompt "build me a game based on this paper, make it interesting and thematic but still conveying key findings" and get a tiny working educational game. (In this case, I used Gemini)
Reposted by ryanbehdad.bsky.social
My top 11 multi-season TV dramas are:
The Wire
Mad Men
Breaking Bad
The Leftovers
Andor
The Americans
Game of Thrones
Succession
Better Call Saul
The Sopranos
Halt and Catch Fire
The Wire
Mad Men
Breaking Bad
The Leftovers
Andor
The Americans
Game of Thrones
Succession
Better Call Saul
The Sopranos
Halt and Catch Fire
I think we need a top ten list. I watched The Americans based on your recommendation and i really enjoyed it. I’m going to start Mad Men next.
May 14, 2025 at 9:17 PM
My top 11 multi-season TV dramas are:
The Wire
Mad Men
Breaking Bad
The Leftovers
Andor
The Americans
Game of Thrones
Succession
Better Call Saul
The Sopranos
Halt and Catch Fire
The Wire
Mad Men
Breaking Bad
The Leftovers
Andor
The Americans
Game of Thrones
Succession
Better Call Saul
The Sopranos
Halt and Catch Fire
Reposted by ryanbehdad.bsky.social
Reposted by ryanbehdad.bsky.social
For academics that are skeptical about AI's ability to provide perceptive & useful results, a fast way to test your assumptions is to give one of your papers as a PDF to o3 (or Gemini 2.5 but it lacks some tools), with the prompt:
"critique this paper"
I have found very good accuracy & insight...
"critique this paper"
I have found very good accuracy & insight...
May 11, 2025 at 1:56 AM
For academics that are skeptical about AI's ability to provide perceptive & useful results, a fast way to test your assumptions is to give one of your papers as a PDF to o3 (or Gemini 2.5 but it lacks some tools), with the prompt:
"critique this paper"
I have found very good accuracy & insight...
"critique this paper"
I have found very good accuracy & insight...
Reposted by ryanbehdad.bsky.social
Being able to see the "thinking" of an AI can occasionally feel infantalizing when you see that it is clearly dumbing things down for you.
May 7, 2025 at 6:46 PM
Being able to see the "thinking" of an AI can occasionally feel infantalizing when you see that it is clearly dumbing things down for you.
Reposted by ryanbehdad.bsky.social
My favorite local model right now is a bit of surprise to me: I'm really enjoying the relatively tiny Qwen3-8B, running the 4bit quantized version on my Mac using MLX
It's surprisingly capable given it's a 4.3GB download and uses just 4-5GB of RAM while running
simonwillison.net/2025/May/2/q...
It's surprisingly capable given it's a 4.3GB download and uses just 4-5GB of RAM while running
simonwillison.net/2025/May/2/q...
Note on 2nd May 2025
Having tried a few of the [Qwen 3 models](https://simonwillison.net/2025/Apr/29/qwen-3/) now my favorite is a bit of a surprise to me: I'm really enjoying [Qwen3-8B](https://huggingface.co/Qwen/Qwen3-...
simonwillison.net
May 2, 2025 at 11:45 PM
My favorite local model right now is a bit of surprise to me: I'm really enjoying the relatively tiny Qwen3-8B, running the 4bit quantized version on my Mac using MLX
It's surprisingly capable given it's a 4.3GB download and uses just 4-5GB of RAM while running
simonwillison.net/2025/May/2/q...
It's surprisingly capable given it's a 4.3GB download and uses just 4-5GB of RAM while running
simonwillison.net/2025/May/2/q...
Reposted by ryanbehdad.bsky.social
How much do you think it costs to make a pair of Nike shoes in Asia?
Before you open this thread, take a guess. Settle on a number and then compare it to what you'll learn in this thread. 🧵
Before you open this thread, take a guess. Settle on a number and then compare it to what you'll learn in this thread. 🧵
April 11, 2025 at 5:19 PM
How much do you think it costs to make a pair of Nike shoes in Asia?
Before you open this thread, take a guess. Settle on a number and then compare it to what you'll learn in this thread. 🧵
Before you open this thread, take a guess. Settle on a number and then compare it to what you'll learn in this thread. 🧵
Reposted by ryanbehdad.bsky.social
Your outie's code works on the first try
February 19, 2025 at 3:51 AM
Your outie's code works on the first try
Reposted by ryanbehdad.bsky.social
absolutely wild. i went to youtube to watch a video and the autoplay ad was an AI fake of tom hanks spouting fake diabetes cures. google is making money off this???
February 13, 2025 at 8:42 PM
absolutely wild. i went to youtube to watch a video and the autoplay ad was an AI fake of tom hanks spouting fake diabetes cures. google is making money off this???
Reposted by ryanbehdad.bsky.social
Ever felt like this?
February 6, 2025 at 9:43 AM
Ever felt like this?
Reposted by ryanbehdad.bsky.social
A year ago, I took a risk & spent quite some time on a ML competition. It paid off—I won 4th place overall & 1st in explainability!
Here's a summary of the journey, challenges, & key insights from my winning solution (water supply forecasting)
Here's a summary of the journey, challenges, & key insights from my winning solution (water supply forecasting)
February 4, 2025 at 1:18 PM
A year ago, I took a risk & spent quite some time on a ML competition. It paid off—I won 4th place overall & 1st in explainability!
Here's a summary of the journey, challenges, & key insights from my winning solution (water supply forecasting)
Here's a summary of the journey, challenges, & key insights from my winning solution (water supply forecasting)
Reposted by ryanbehdad.bsky.social
The raw chain of thought from DeepSeek is fascinating, really reads like a human thinking out loud. Charming and strange.
January 20, 2025 at 7:26 PM
The raw chain of thought from DeepSeek is fascinating, really reads like a human thinking out loud. Charming and strange.
Reposted by ryanbehdad.bsky.social
> Nobody tells you the variables you should be regressing. What's the target? What's the source? Do you notice when results are rubbish? ... That's why I think you need smart people who appear to do something technically easy but actually not so easy.
news.ycombinator.com/item?id=1906...
news.ycombinator.com/item?id=1906...
!["...I joined a hedged fund, Renaissance Technologies, I'll make a comment about that. It's funny that I think the most important thing to do on data analysis is to do the simple things right. So, here's a kind of non-secret about what we did at renaissance: in my opinion, our most important statistical tool was simple regression with one target and one independent variable. It's the simplest statistical model you can imagine. Any reasonably smart high school student could do it. Now we have some of the smartest people around, working in our hedge fund, we have string theorists we recruited from Harvard, and they're doing simple regression. Is this stupid and pointless? Should we be hiring stupider people and paying them less? And the answer is no. And the reason is nobody tells you what the variables you should be regressing [are]. What's the target. Should you do a nonlinear transform before you regress? What's the source? Should you clean your data? Do you notice when your results are obviously rubbish? And so on. And the smarter you are the less likely you are to make a stupid mistake. And that's why I think you often need smart people who appear to be doing something technically very easy, but actually usually not so easy.]
[[at] my hedge fund, which was not a very big company, we had 7 Phd's just cleaning data and organizing the databases]"](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:ilhs5ksfpze2l5rqgig4ycl6/bafkreiblk7ckawi7cdgnaven5ulbt767s7gs56nreoa7s4cdsjj35jdoo4@jpeg)
January 17, 2025 at 1:22 AM
> Nobody tells you the variables you should be regressing. What's the target? What's the source? Do you notice when results are rubbish? ... That's why I think you need smart people who appear to do something technically easy but actually not so easy.
news.ycombinator.com/item?id=1906...
news.ycombinator.com/item?id=1906...
Reposted by ryanbehdad.bsky.social
I don't think people know that making plausible fake videos is now something that doesn't require a lot of skill or time
I made this Veo video of a man trying out a flying skateboard at CES from a text prompt, 2nd try. Notice the shadows, green light, his body shifts for balance
I made this Veo video of a man trying out a flying skateboard at CES from a text prompt, 2nd try. Notice the shadows, green light, his body shifts for balance
January 14, 2025 at 2:21 AM
I don't think people know that making plausible fake videos is now something that doesn't require a lot of skill or time
I made this Veo video of a man trying out a flying skateboard at CES from a text prompt, 2nd try. Notice the shadows, green light, his body shifts for balance
I made this Veo video of a man trying out a flying skateboard at CES from a text prompt, 2nd try. Notice the shadows, green light, his body shifts for balance
Reposted by ryanbehdad.bsky.social
This is neat. I'm playing with Virat Singh's open source AI "hedge fund" where different "agents" each performs a different analysis and gives them to a portfolio manager agent to make final calls. Also does back-testing to see how well it performs.
.
Early days (don't trade on it), but interesting
.
Early days (don't trade on it), but interesting
January 10, 2025 at 7:23 PM
This is neat. I'm playing with Virat Singh's open source AI "hedge fund" where different "agents" each performs a different analysis and gives them to a portfolio manager agent to make final calls. Also does back-testing to see how well it performs.
.
Early days (don't trade on it), but interesting
.
Early days (don't trade on it), but interesting
Reposted by ryanbehdad.bsky.social
What percentage of students are you willing to falsely accuse of cheating with AI?
There is a trade-off between false accusations and detection rates for AI. At a 10% false positive rate, detectors find 80% or less of AI content. At a 1% rate most find 60% or less.
Don’t trust AI detectors!
There is a trade-off between false accusations and detection rates for AI. At a 10% false positive rate, detectors find 80% or less of AI content. At a 1% rate most find 60% or less.
Don’t trust AI detectors!
January 6, 2025 at 2:30 PM
What percentage of students are you willing to falsely accuse of cheating with AI?
There is a trade-off between false accusations and detection rates for AI. At a 10% false positive rate, detectors find 80% or less of AI content. At a 1% rate most find 60% or less.
Don’t trust AI detectors!
There is a trade-off between false accusations and detection rates for AI. At a 10% false positive rate, detectors find 80% or less of AI content. At a 1% rate most find 60% or less.
Don’t trust AI detectors!
Reposted by ryanbehdad.bsky.social
If you have ever tried to read free books from sites like Project Gutenberg, you noticed that they can be uncomfortable to read, due to their layouts, type & occasional errors
This project takes those free books and makes them beautiful (and still free). standardebooks.org
This project takes those free books and makes them beautiful (and still free). standardebooks.org
December 25, 2024 at 1:45 PM
If you have ever tried to read free books from sites like Project Gutenberg, you noticed that they can be uncomfortable to read, due to their layouts, type & occasional errors
This project takes those free books and makes them beautiful (and still free). standardebooks.org
This project takes those free books and makes them beautiful (and still free). standardebooks.org
Reposted by ryanbehdad.bsky.social
AI Engingeering Primer
This should be a good starting point. A little bit of advice, and then a lot of topics to cover and l places to get information
timkellogg.me/blog/2024/12...
This should be a good starting point. A little bit of advice, and then a lot of topics to cover and l places to get information
timkellogg.me/blog/2024/12...
AI Engineering Primer - Tim Kellogg
timkellogg.me
December 20, 2024 at 12:44 PM
AI Engingeering Primer
This should be a good starting point. A little bit of advice, and then a lot of topics to cover and l places to get information
timkellogg.me/blog/2024/12...
This should be a good starting point. A little bit of advice, and then a lot of topics to cover and l places to get information
timkellogg.me/blog/2024/12...
Reposted by ryanbehdad.bsky.social
This is pretty damn nifty! @hamel.bsky.social @projectjupyter.bsky.social
#datascience #jupyternotebooks
www.answer.ai/posts/2024-1...
#datascience #jupyternotebooks
www.answer.ai/posts/2024-1...
nbsanity - Share Notebooks as Polished Web Pages in Seconds – Answer.AI
Transform your GitHub Jupyter notebooks into beautiful, readable web pages with a single URL change. No setup required.
www.answer.ai
December 20, 2024 at 3:33 PM
This is pretty damn nifty! @hamel.bsky.social @projectjupyter.bsky.social
#datascience #jupyternotebooks
www.answer.ai/posts/2024-1...
#datascience #jupyternotebooks
www.answer.ai/posts/2024-1...
Reposted by ryanbehdad.bsky.social
I tell you what, that Santa Claus has got real presence about him
December 18, 2024 at 5:32 PM
I tell you what, that Santa Claus has got real presence about him
Reposted by ryanbehdad.bsky.social
Google has a knack for making non-chatbot interfaces for serious work with LLMs.
When I demo them, both NotebookLM & Deep Research are instantly understandable and fill real organizational needs. They represent a tiny range of AI capability, but they are easy for everyone to get and use right away.
When I demo them, both NotebookLM & Deep Research are instantly understandable and fill real organizational needs. They represent a tiny range of AI capability, but they are easy for everyone to get and use right away.
December 14, 2024 at 3:11 AM
Google has a knack for making non-chatbot interfaces for serious work with LLMs.
When I demo them, both NotebookLM & Deep Research are instantly understandable and fill real organizational needs. They represent a tiny range of AI capability, but they are easy for everyone to get and use right away.
When I demo them, both NotebookLM & Deep Research are instantly understandable and fill real organizational needs. They represent a tiny range of AI capability, but they are easy for everyone to get and use right away.
Reposted by ryanbehdad.bsky.social
Revenge is a dish best served after 14 years
December 7, 2024 at 7:36 PM
Revenge is a dish best served after 14 years