Apple Vision Pro customers face a 25-minute in-store sales pitch (3 minute read)
The Apple Vision Pro will be available from February 2 starting from $3,499. Customers waiting to buy an Apple Vision Pro may have to sit through a lengthy sales pitch that includes a 25-minute in-store demonstration on how to use the headset. More than a dozen demo units will be available at any one time at some of Apple's largest stores to accommodate the appointments. Customers will be able to purchase the Apple Vision Pro online without visiting a store to try the headset out.
|
How Threads will integrate with the Fediverse (38 minute read)
The Fediverse consists of small social networks that can be hosted by an individual or company whose users can communicate and interact with users on other similar networks. This article details how Meta plans to integrate Threads with the Fediverse. Meta engaging positively with its user base is difficult due to its size. Integrating with the Fediverse will let Meta step back and focus instead on engaging with organizations that represent the interests of the Fediverse.
|
|
Science & Futuristic Technology
|
Some highlights on the Starship side of the All Hands (2 minute read)
Starship's second flight almost made it to orbit - if it had a payload, that payload would have made it to orbit. SpaceX thinks it has a good shot at reaching orbit with the third flight. It will then focus on achieving full and rapid reusability. There are two towers at Starbase so one can be upgraded while the other is used for launches.
|
New Anthropic Paper: Sleeper Agents (3 minute read)
Anthropic recently trained large language models to act secretly malicious. It found that despite best efforts at alignment training, deception still slipped through. The alignment training only made models look safe. Anthropic's research shows how standard safety training might not actually ensure safety and might end up giving people a false sense of security.
|
|
Programming, Design & Data Science
|
Azure Security Best Practices (Sponsor)
Security is a key priority for any CISO of any organization with an extensive footprint in Azure. This cheat sheet provides actionable recommendations that can help you strengthen your Azure cloud security posture. We’ll explore detailed aspects of Azure best practices, from role-based access control (RBAC) to cloud security posture management, that you can adapt to secure your Azure subscriptions. Get the FREE Azure Cheat Sheet
|
Hidden Changes in GPT-4, Uncovered (14 minute read)
Recent changes made to ChatGPT's GPT-4 interface significantly altered the model's web browsing tools. The changes prevent the model from directly citing quotes from webpages and limits viewing of full content. This article analyzes the underlying mechanism of GPT-4 web browsing, the hidden changes, and their implications. OpenAI is increasingly under scrutiny regarding copyright laws and the reproduction of content on the internet. These changes seem to be aimed at eliminating the issue.
|
remoteStorage (GitHub Repo)
remoteStorage combines the localStorage API with a remote server to persist data across browsers and devices. It allows apps to easily read and write data on the fly while maintaining state across browsers and devices. remoteStorage features a simple API and it works with all JavaScript frameworks.
|
|
Instagram’s co-founders are shutting down their Artifact news app (1 minute read)
Artifact, a news app created by Instagram's co-founders, is shutting down a year after launch. The app used an AI-driven approach to suggest news users might like to read. The company decided that the market opportunity for the app wasn't big enough to warrant continued investment. Artifact will still let users read news until the end of February, but users can no longer add new comments or posts.
|
My AI Timelines Have Sped Up (Again) (30 minute read)
This article is an update to a post in 2020 that made predictions on when artificial general intelligence (AGI) will be achieved. It defines AGI as an AI system that matches or exceeds humans at almost all economically valuable work. The previous post predicted a 10% chance by 2035, but this has now changed to a 10% chance by 2028 and a 25% chance by 2035. The author believes there is a 90% chance that humans will develop AGI by 2070.
|
|
marimo (GitHub Repo)
marimo is a reactive notebook for Python that allows users to rapidly experiment with data and models, code with confidence, and productionize notebooks as pipelines or interactive web apps.
|
|
Want the best of TLDR? 🏆
Refer a friend to TLDR using the referral link below, and we will send you the TLDR Hall of Fame, our 50 best stories of all time!
Your Referral Link -
https://tldr.tech/tech?ref=2775316
|
We help cutting edge companies hire world class technical talent through our job listings. If you're hiring software engineers, AI/ML engineers, product managers, designers or other tech talent, click here to learn more.
|
If your company is interested in reaching an audience of tech executives, decision-makers and engineers, you may want to advertise with us.
If you have any comments or feedback, just respond to this email!
Thanks for reading,
Dan Ni and Stephen Flanders
|
|
|
|