Today my daughter and I are pleased to announce the release of our first open source project together!
github.com/humphrem/action
ACTION is an AI detector pipeline for automatically extracting animal/fish clips from camera trap videos. It uses the YOLO-Fish and MegaDetector models
No Margot Robbie in a bathtub today. Just a hedgehog who's been watching private credit for months and needs you to read this carefully.
🦔The biggest banks on Wall Street are launching a new tool next week to bet against private credit funds, the same $3 trillion corner of finance that quietly holds money from pension funds and retirement accounts. JPMorgan, Bank of America, Barclays, Deutsche Bank, and Goldman Sachs are working with S&P Global to launch a credit default swap index called CDX Financials, allowing investors to profit if private credit fund managers including Apollo, Ares, and Blackstone run into trouble.
Credit default swaps are insurance contracts that pay out when borrowers default. Banks want protection against their own exposure. Hedge funds want to profit from a downturn. This comes as Blue Owl reported 41% redemption requests last quarter and Carlyle's fund was hit with a 15.7% redemption request this week, more than three times its normal limit.
My Take
I want to explain why this matters for regular people because the language around it is designed to be confusing.
Private credit is a market where investment firms lend directly to companies outside the traditional banking system. Millions of Americans have indirect exposure through pension funds and 401k plans without knowing it. When those loans go bad and investors try to get their money out simultaneously, funds gate withdrawals, meaning you simply cannot access your money when you need it.
What the banks are doing now is building the infrastructure to profit when that happens. Credit default swap indexes were central to the 2008 financial crisis, and that doesn't mean we're heading there again. But when the largest banks on Wall Street simultaneously decide they need protection against private credit losses, and hedge funds are lining up to bet on a collapse, people with retirement savings in these funds deserve to understand what the people on the other side of that trade are seeing.
Hedgie🤗
I figured this out, it's my mistake: the provisioning key is admin level, and my account in the org is "member" level, so it shows me all the API keys I own.
Human Creativity in the Age of LLMs
Randomized Experiments on Divergent and Convergent Thinking
"Our findings reveal that while LLM assistance can provide short-term boosts in creativity during assisted tasks, it may inadvertently hinder independent creative
performance when users work without assistance, raising concerns about the long-term impact on human creativity and cognition."
by @1harsh_kumar J. Vincentius, E. Jordan, @ashton1anderson
📖 arxiv.org/pdf/2410.03703
An incredible story, told v well! Nice to see mention of @inaturalist, the best source for up to date distribution records of species. With so much intentional & continued spreading of the species by people its range here won’t attract scientific study! 🦎
inaturalist.org/observations…
I spend more time fixing my code for eslint to continue working in my projects than I do fixing things that eslint finds. I'm getting to the point where I feel like it's not worth the extra maintenance.
How can we make our documentation, notes, and libraries more LLM-friendly? Create a single .txt file. 🤔 llms.txt proposed by @jeremyphoward, now adopted by @AnthropicAI proposes as a new standard to provide a structured, concise overview a single markdown file with key information to make the content more accessible and usable for AI agents. 🧠
TL;DR:
📄 Add a new standard file /llms.txt at website root, similar to robots.txt but specifically for LLMs
🎯 Use Markdown format for both human and LLM readability while maintaining strict structural requirements
🔄 Suggests appending .md to URLs to provide clean markdown versions of regular web pages
💡 Optional sections to handle context window limitations effectively
🛠️ created llms-txt to parse llms.txt for better use
🌐 Compatible with existing web standards like robots.txt and sitemap.xml
Friday docs feature drop:
You can now access all of our docs concatenated as a single plain text file that can be fed in to any LLM.
Here's the url route: docs.anthropic.com/llms-full…
Honestly, at this point If you give me a programming interview and don't let me use AI assistance you won't get a very realistic idea of what I'm actually capable of
Coding interviews should be using ai. If you disagree you literally don’t understand what ai is for
Sure you can use it to fill gaps & find easy answers
More importantly you can use it to accelerate and do much bigger things
Failing to select for this skill will ruin tech co’s
Today we are releasing #WebVM 2.0: a complete Linux Desktop Environment running in the browser. This incredible achievement of my team @leaningtech is possible thanks for #CheerpX an x86 VM in #WebAssembly.
We'll keep pushing the limits, stay tuned.
labs.leaningtech.com/blog/we…
Hurl is used in the official Python repo 🤯!
Hurl tests all the redirection of Python docs; as one the PR says «We have ~9,400 test cases for redirects [...] Hurl runs these in parallel, so all tests only take a second or so to run.»
github.com/python/psf-salt/t…
Pleased to announce Anchoreum, my free coding game for learning CSS anchor positioning. In the same spirit as Flexbox Froggy and Grid Garden. Chrome and Edge only for now. anchoreum.com