TLDR

TLDR 2023-07-21

6 sections16 entriesinline urlcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01GENERAL

FERN - INSTANTLY CREATE A STRIPE-LIKE API EXPERIENCE (GITHUB REPO SPONSOR)

Fern is an open source tool to automatically generate idiomatic client libraries and Stripe-like API documentation. Find out why companies like Scale AI and Flatfile already use Fern to develop their APIs and generate their SDKs, learn more and star us on GitHub [https://danni763618.typeform.com/to/c4o3Nwrq]!

danni763618.typeform.comGitHub repository

02BIG TECH & STARTUPS

CHATGPT CAN NOW REMEMBER WHO YOU ARE AND WHAT YOU WANT

ChatGPT's new custom instructions feature lets users tell the chatbot things it should always know about them and how to answer questions. It is designed to make ChatGPT quicker and more helpful as a virtual assistant. Custom Instructions is available now for ChatGPT Plus subscribers everywhere but the UK and EU on an opt-in basis. The feature is still in beta. It works everywhere that ChatGPT does.

theverge.com3 min read

AZURE CUSTOMERS

Microsoft is making Meta's Llama 2 model available on its Azure cloud computing service. Meta said in a blog post that Microsoft is its preferred partner for Llama 2. The partnership will give developers more choice in the models they build on. Meta says that it is not focused on monetization for Llama 2. Llama 2 will also be available through Amazon web services and Hugging Face.

cnbc.com2 min read

03SCIENCE & FUTURISTIC TECHNOLOGY

SYSTEM

The US Federal Reserve is set to launch FedNow, a service that will allow US citizens to send and receive funds in seconds. FedNow has been in the works since 2019. It will launch with 41 banks and 15 service providers, with more banks and credit unions to be onboarded later this year. FedNow payments will settle directly in central bank accounts. There will be no fees from FedNow to use the service. FedNow has a maximum payment limit of $500,000, but banks can choose to lower that cap if required.

reuters.com2 min read

TEACHING ROBOTS TO TEACH OTHER ROBOTS

Lifelong Learning (LL) is a growing area of machine learning research in which AI agents continually learn as they encounter new tasks while retaining knowledge of previous tasks. A team of LL researchers recently discovered that AI agents can quickly learn a wide range of tasks by sharing knowledge with each other at the same time. Their system has many similarities with the concept of crowdsourcing. The researchers believe that their technique could transform any field that requires vast, diverse knowledge or dealing with complex systems.

viterbischool.usc.edu4 min read

04PROGRAMMING, DESIGN & DATA SCIENCE

Sponsored

QA Wolf

qawolf.comSponsor

GITHUB COPILOT CHAT BETA NOW AVAILABLE FOR EVERY ORGANIZATION

A limited public beta release of GitHub Copilot Chat is now available to enterprise companies and organizations on Visual Studio and VS Code. GitHub Copilot Chat turns GitHub Pilot into a context-aware conversational assistant, allowing developers to execute tasks with prompts. The feature will help developers create much faster and democratize software development as we know it. Information on how organization admins can enable the feature is available in the article.

github.blog4 min read

OLLAMA

Ollama allows users to run and package large language models on macOS. It includes a library of open-source models, including Llama2, Orca Mini, Vicuna, and Nous-Hermes. At least 8 GB of RAM is needed to run the 3B models, 16 GB for the 7B models, and 32 GB for the 13B models.

github.comGitHub repository

05MISCELLANEOUS

CAN'T AFFORD

Netflix is in better financial shape than its industry rivals and can make better deals. It will likely not be affected by ongoing strikes in Hollywood by writers and actors. Netflix could force a settlement to the strikes before its rivals get back to business. Wall Street analysts estimate that Netflix will be earning up to $9.5 billion annually by 2025.

cnbc.com10 min read

REQUEST

Several companies, including Microsoft, Google, and OpenAI, have committed to certain safeguards for their technology until Congress passes legislation to regulate AI. Officials want to make sure that the technology is developed responsibly and in a way that benefits society without negatively impacting the safety, rights, and democratic values of the public. The tech firms have agreed to eight measures, including letting independent experts test models for bad behavior, investing in cybersecurity, flagging societal risks, and watermarking AI-generated audio and visual content. A full list of the measures is available in the article.

engadget.com2 min read