3751
TypeChat
July 20, 2023 by Anders Hejlsberg, Steve Lucco, Daniel Rosenwasser, Pierce Boggan, Umesh Madan, Mike Hopcroft, and Gayathri Chandrasekaran
New
The latest blog posts to take off on Hacker News. 6,334 posts.
3752
A framework to securely use LLMs in companies – Part 1: Overview of Risks
Part 1 of a multi-part series on using LLMs securely within your organisation. This post provides a framework to categorize risks based on different use cases and deployment type.
3753
Use pytorch2+cu118 with ADA hardware for 50%+ speedup
GPUX is a distributed supercomputer.
3754
SHOW-1 and Showrunner Agents in Multi-Agent Simulations
South Park AI is an example of how to generate high-quality episodic content through the use of showrunner agents in the simulation. #AI #simulation #multi-agent
3755
3756
Accessing Llama 2 from the command-line with the LLM-replicate plugin
The big news today is Llama 2, the new openly licensed Large Language Model from Meta AI. It’s a really big deal: Llama 2 is available for free for research …
3757
Generative AI space and the mental imagery of alien minds
AI and the ruliad in the Wolfram Physics Project provide a new way to consider how alien minds might perceive the world. Stephen Wolfram explains how this artificial neuroscience experiment works.
3758
fMRI-to-image with contrastive learning and diffusion priors
a cutting-edge method for reconstructing and retrieving images from fMRI brain activity. Discover how fMRI enables detailed analysis of brain functions and aids in evaluating brain treatments. MindEye's training on the Natural Scenes Dataset brings new insights to neuroscience research. Explore the
3759
Bad numbers in the “gzip beats BERT” paper?
The recent paper, “Low-Resource” Text Classification: A Parameter-Free Classification Method with Compressors by Jiang et al. [ link ] recently received a lot of attention on twitter.
3760
How to Use AI to Do Stuff: An Opinionated Guide
Covering the state of play as of Summer, 2023
3761
WormGPT – The Generative AI Tool Cybercriminals Are Using
Protect sensitive data from phishing and social engineering with Varonis Interceptor, AI-native email security with 99.9% detection and real-time response.
3762
The Problem with LangChain
LangChain is complicated, so it must be better. Right?
3763
Faster neural networks straight from JPEG (2018)
Uber AI Labs introduces a method for making neural networks that process images faster and more accurately by leveraging JPEG representations.
3764
Bard’s latest update: more features, languages and countries
Today we're expanding Bard to new languages and countries, and launching new features to help you get creative and boost your productivity.
3765
NotebookLM: An AI Notebook
We’re rolling out NotebookLM, an experimental offering from Google Labs to summarize information, complex ideas and brainstorm new connections.
3766
20x faster than pgvector: HNSW index in Postgres with pg_embedding
We’re excited to announce the release of our pg_embedding extension for Postgres and LangChain! The new pg_embedding extension brings 20x the speed for 99% accuracy to graph-based approximate nearest neighbor search to your Postgres databases. While the pgvector extension with IVFFlat indexing has been a popular…
3767
3768
3769
AI Safety and the Age of Dislightenment
Model licensing & surveillance will likely be counterproductive by concentrating power in unsustainable ways
3770
Claude 2
Claude 2 brings improved performance, a 100K token context window, and 2x better harmlessness, available via API and the claude.ai beta.
3771
Introduction to vector similarity search (2022)
Learn what vector search is and the metrics pertinent to decide the distance (or similarity) between objects.
3772
PoisonGPT: We hid a lobotomized LLM on Hugging Face to spread fake news
We will show in this article how one can surgically modify an open-source model, GPT-J-6B, and upload it to Hugging Face to make it spread misinformation while being undetected by standard benchmarks.
3773
Machine Unlearning Challenge
Posted by Fabian Pedregosa and Eleni Triantafillou, Research Scientists, Google Deep learning has recently driven tremendous progress in a wide arr...
3774
Train an AI model once and deploy on any cloud
Organizations are increasingly adopting hybrid and multi-cloud strategies to access the latest compute resources, consistently support worldwide customers, and optimize cost. However…
3775
Google to explore alternatives to robots.txt
We’re kicking off a public discussion across the web and AI communities to develop new machine-readable means to provide web publisher choice and control.
3776
3777
GPT-4 API General Availability
GPT-3.5 Turbo, DALL·E and Whisper APIs are also generally available, and we are releasing a deprecation plan for older models of the Completions API, which will retire at the beginning of 2024.
3778
Demystifying Text Data with the Unstructured Python Library
In the world of data, textual data stands out as being particularly complex. It doesn’t fall into neat rows and columns like numerical data does. As a side project, I’m in the process of developing my own personal AI assistant. The objective is to use the data within my notes and documents to answer my questions. The…
3779
The many ways that digital minds can know – A better way to think about LLMs
LLMs do all the things people say they do at once, even if those people vociferously disagree, and they’re all useful.
3780
Introducing Superalignment
We need scientific and technical breakthroughs to steer and control AI systems much smarter than us. To solve this problem within four years, we’re starting a new team, co-led by Ilya Sutskever and Jan Leike, and dedicating 20% of the compute we’ve secured to date to this effort. We’re looking for excellent ML…