theaspiringnerd.com - Profile | ThreadSky | a Reddit-style client for Bluesky

Deep Research is available to ChatGPT Plus users

submitted 1 day ago • 1 comment

Equally funny and terrifying!

submitted 3 days ago • 0 comments

Privacy advocates have long resisted attempts by governments and corporations to collect and store citizens' personal data. This is the reason. To prevent authoritarian, unaccountable, secret police from knowing everything about our lives and using our data against us.

submitted 8 days ago • 11 comments

a friend asked "how do I add a directory to my PATH" and I couldn't find any directions I liked so I wrote up a quick guide: jvns.ca/blog/2025/02...

submitted 13 days ago • 11 comments

"Claude the Eagles won the superbowl, do something to celebrate, make it amazing." (boringly sane) "Make it more Philly" (adds its attempts at Philly slang, cheese steak, Liberty Bell) "Even more Philly!" (adds greased poles, batteries being thrown at Santa who is being boo'ed, Gritty, etc)

submitted 16 days ago • 1 comment

This was such a good listen. Make sure you stay till the end when Doctorow sums up the crucial difference between hope and optimism.

submitted 19 days ago • 0 comments

Every 4 years, a team of libraries & research organizations work together to preserve material from U.S. government websites during the transition of administrations. 🗳️ Get the latest on the 2024/2025 End of Term Web Archive @eotarchive.org ➡️ blog.archive.org/2025/02/06/u...

submitted 20 days ago • 32 comments

As someone trained as a trade economist, it is my duty to share the 1929-1933 Kindleberger Spiral, showing the month-month decline in global trade due to the combined factors of the (global) Great Depression and retaliatory tariffs. Smooth Hawley is implemented mid-June 1930.

submitted 25 days ago • 51 comments

Genuinely unsure how OpenAI survives. GenAI doesn't have a killer app. It assumed you needed hundreds of billions of capex, which turned out to be untrue. OpenAI's left with using DeepSeek's tech, which is an admittance of defeat, or continuing with its own expensive and uncompetitive models.

submitted 25 days ago • 8 comments

Here's a fun prompt injection challenge: can you get DeepSeek R1 running on chat.deepseek.com to leak its system prompt? I'm finding it's pretty robust at reasoning about how it shouldn't do that

submitted 32 days ago • 8 comments

No matter how much you fight it, I find that the visible chain-of-thought from DeepSeek makes it nearly impossible to avoid anthropomorphizing the thing The visible first-person "thinking" makes you feel like you are reading a diary of a somewhat tortured soul who wants to help. Weird interaction

submitted 35 days ago • 12 comments

Newsletter: No, Trump didn’t make $50 billion from his memecoin. www.citationneeded.news/trump-memeco...

submitted 37 days ago • 29 comments

to recap: — Trump was for the TikTok ban — which the GOP Congress passed — and the GOP Supreme Court upheld — and even though Biden said he wouldn't enforce the ban, TikTok shut down anyway — so that Trump could "save" it — and the media fell for it we really do live in the dumbest fucking country

submitted 38 days ago • 616 comments

I'd been waiting for someone to write something practical on using o1 and I think this piece by Ben Hylak nails it. (Obviously do read the OpenAI docs and so on, but o1 is something you have to really treat differently to get the most out of.) www.latent.space/p/o1-skill-i...

submitted 45 days ago • 1 comment

American teens are flocking to a new video app

submitted 42 days ago • 617 comments

My notes on the new OpenAI ChatGPT scheduled tasks feature, including a copy of the new tool instructions from a leaked system prompt simonwillison.net/2025/Jan/15/...

submitted 42 days ago • 9 comments

If you are an academic, it can be instructive to work on a paper with AI. Pretend you are working with a grad student & see what happens. Generally o1 is best for well-defined heavy intellectual tasks, Gemini for synthesizing lots of text, and Claude for writing & theorizing. This varies by field.

submitted 45 days ago • 8 comments

My 8000-word note on agents: huyenchip.com//2025/01/07/... 1. An AI-powered agent's capability is determined by its tools and its planning ability 2. How to select the best tools for your agent 3. How to augment a model’s planning capability 4. Agent’s failure modes Feedback is much appreciated!

submitted 50 days ago • 4 comments

Risky post... here are my 1, 3 and 6 year predictions for LLMs/AI, expanded from my appearance on the Oxide and Friends podcast (My confidence in my ability to predict the future is extremely low) simonwillison.net/2025/Jan/10/...

submitted 47 days ago • 9 comments

For those of us Product Managers who are a little less tech-savvy than software engineers, here’s a short blog post I wrote about how I turned an idea into a working app in just a few hours of tinkering with Claude and ChatGPT. #llm #chatgpt #claude #productmanagement

submitted 51 days ago • 1 comment

The Breachies 2024 highlights the most significant data breaches of the year, focusing on those with novel information, notable security flaws, or massive impact. www.eff.org/deeplinks/20...

submitted 52 days ago • 0 comments

Here's why "alignment research" when it comes to LLMs is a big mess, as I see it. Claude is not a real guy. Claude is a character in the stories that an LLM has been programmed to write. Just to give it a distinct name, let's call the LLM "the Shoggoth".

submitted 69 days ago • 9 comments

New year, new blog post: I had a random question, what happens when LLMs are prompted to write better code, again and again? Do they actually write better code? The answer is yes*! minimaxir.com/2025/01/writ...

submitted 55 days ago • 8 comments

Every year on 1 January, Public Domain Day celebrates the artistic works that enter the public domain, becoming free for everyone to use, share, and adapt. Wikimedia volunteers are at the forefront of sharing these works while navigating copyright restrictions. ➡️ w.wiki/CbAe

submitted 56 days ago • 2 comments

I think everyone who has an opnion, good or bad, about LLMs, should read how @simonwillison.net has summer up what’s happened in the space this year. He’s the most credible, most independent, most honest, and most technically fluent person watching the space. simonwillison.net/2024/Dec/31/...

submitted 57 days ago • 13 comments

Eight Clams Control This Polish City’s Water Supply. “If the waters are clean, these [cyborg] mussels stay open and happy. But when water quality drops too low, they close off and shut the water supply of millions of people with them.” [kottke.org]

submitted 58 days ago • 4 comments

Zoomers: "Hey why these mf so good at command line interfaces" Millennials: "You wouldn't last an hour in the asylum where they raised me"

submitted 58 days ago • 21 comments

All this dystopian x-risk AGI futurist BS is (1) an expression of male technological ego and (2) a distraction from present-tense crisis. Fascism is upon us, folks. Stop with your p-doom doomer bullshit and pay attention to now.

submitted 59 days ago • 12 comments

This year, we worked swiftly to save legacy media sites Vice.com and MTVNews before decades worth of valuable journalism could be erased. These sites are now searchable on the Wayback Machine! Help us in saving these resources:: https://archive.org/donate/?origin=blsky-eoy2024

submitted 60 days ago • 56 comments

I’ve been doing that for years too! At some point, I turned it into a weekly email wrap-up but couldn’t keep up with the pace. I’m back at it whenever I can now.

submitted 66 days ago • 0 comments

New AI Snake Oil essay: Last month the AI industry's narrative suddenly flipped — model scaling is dead, but "inference scaling" is taking over. This has left people outside AI confused. What changed? Is AI capability progress slowing? We look at the evidence. 🧵 www.aisnakeoil.com/p/is-ai-prog...

submitted 69 days ago • 3 comments

Today's post by @vale.rocks is one of my favourites because it's fun and interesting, and I learned something new! htmhell.dev/adventcalend...

submitted 68 days ago • 2 comments

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/...

submitted 78 days ago • 10 comments

Canvas and Projects might be what was missing for me to pick ChatGPT over Claude as my daily tool.

submitted 75 days ago • 0 comments

Incredible project!

submitted 76 days ago • 0 comments

A terrific essay about how and why we should understand AI as a political project injected into our technology

submitted 79 days ago • 5 comments

Tired of wrestling with complex RAG pipelines? 🚀 Enter Roaming RAG: a simpler way to make your LLMs find answers in well-structured docs. No vector databases, no headaches—just rich, structured context. 👉 Read how it works: arcturus-labs.com/blog/2024/11...

submitted 82 days ago • 3 comments

Product Managers: if you’ve used both Claude and ChatGPT, I’m curious—do you have a favourite? And why? I’m currently paying for ChatGPT Plus because I use custom GPTs, but I’m questioning if it’s worth the money. #productmanagement

submitted 84 days ago • 1 comment

Thanks to LLMs, I no longer have to ever relearn regex. That alone is worth $20/month.

submitted 88 days ago • 46 comments