4291
Being “Confidently Wrong” is holding AI back
The failure mode that stalls “AI for data” efforts or "AI on my APIs" efforts isn’t psychedelic hallucination—it’s confident inaccuracy: plausible answers that are wrong in subtle and costly ways.
Top posts
Every blog post that took off on Hacker News, ranked by the votes of its best submission. 6,334 posts.
4292
Spending Too Much Money on a Coding Agent
On making use of large thinking models.
4293
Devin is now generally available
The work Devin has been doing and how your team can start using Devin today.
4294
A Step Behind the Bleeding Edge: A Philosophy on AI in Dev
This is a memo I published internally to my team at Monarch. I’m sharing it more publicly in case it helps other software engineering teams that are managing the crazy times we’re exper…
4295
4296
Algorithms Are Now Commodities
When I first started writing software, developers had to implement most of the algorithms they used; yes, hardware vendors provided libraries, but the culture was one of self-reliance (except for maths functions, which were technical and complicated).
4297
“The Book of Why” by Pearl and Mackenzie
Judea Pearl and Dana Mackenzie sent me a copy of their new book, “The book of why: The new science of cause and effect.”
4298
The Code-Only Agent
When code execution really is all you need. Exploring what happens when an AI agent can only write and run code.
4299
Structured outputs create false confidence
Constrained decoding seems like the greatest thing since sliced bread, but it often forces models to prioritize output conformance over output quality.
4300
4301
How to Conquer Tensorphobia
A professor at Stanford once said, If you really want to impress your friends and confound your enemies, you can invoke tensor products… People run in terror from the $ \otimes$ symbol. He was explaining some aspects of multidimensional Fourier transforms, but this comment is only half in jest; people get confused by…
4302
Document poisoning in RAG systems: How attackers corrupt AI's sources
RAG poisoning attack walkthrough: an attacker injects fabricated documents into a ChromaDB knowledge base and the LLM reports false financials as fact. Fully reproducible, 100% local, no GPU required.
4303
Speculative KV coding: losslessly compressing KV cache by up to ~4×
Lossless compression of a target model's KV cache by up to 4×, using a cheaper predictor model to drive an arithmetic coder.
4304
The efficient frontier of LLM inference
Inference techniques either move a deployment along the latency–throughput frontier or push the entire frontier out, creating more efficiency to allocate.
4305
Using GPT-4 to measure the passage of time in fiction
Large language models are valuable research assistants, especially when they refuse to follow instructions.
4306
4307
ChatGPT plugins now support Postgres and Supabase
Supabase recently contributed to the OpenAI Retrieval Plugin repo with a Postgres and a Supabase implementation to help developers build ChatGPT plugins using pgvector.
4308
4309
Vibration-minimizing motion retargeting for robotic characters
Creating animations for robotic characters is very challenging due to the constraints imposed by their physical nature. In particular, the combination of fast motions and unavoidable structural deformations leads to mechanical oscillations that negatively affect their performances. Our goal is to automatically…
4310
Running Open-Source AI Models Locally with Ruby
Run AI models locally in Ruby! 💻 Use Ollama to download & run open-source models like Mistral. Integrate them into your Ruby apps using HTTP requests. Great for data security & privacy.
4311
Debugging with intelligence via probabilistic inference
Debugging with intelligence via probabilistic inference Xu et al., ICSE’18 Xu et al. have built a automated debugger that can take a single failing test execution, and with minimal interactio…
4312
What Meta got right with Muse
Design matters more than ever in consumer AI, as long as you have money to play.
4313
Replika AI: Your Money or Your Wife
Isn't there somebody you forgot to ask?
4314
4315
4316
4317
Claude for Creative Work
New connectors let Claude work inside Adobe Creative Cloud, Blender, Autodesk Fusion, SketchUp, and music production tools to support creative workflows.
4318
Role-playing with AI will be a powerful tool for writers and educators
Or, how well can GPT-4 simulate an acid trip in 1963?
4319
Rethinking legal education in the AI era
OverviewArtificial Intelligence is already impacting higher education and the legal profession, and the pace of change appears only to be accelerating. It is thus critical for us to use this moment to carefully reflect on how legal education should adapt. This memo describes the approach we are taking to legal…
4320
Natural language is an unnatural interface
How can we use familiar design patterns to constrain the unwieldy state space of large language models?