More from From ChatGPT to Gemini: how AI is rewriting the internet
Some YouTube Premium subscribers can now jump to the most-watched part of a video, only in the YouTube app, by double-tapping the right side of the screen (which normally skips ahead 10 seconds), then tapping a “Jump ahead” button that appears, according to 9to5Google.
To see if you have the feature and enable it, go to Settings > Try experimental new features.
If you’ve already read our review of the Rabbit R1 but haven’t gotten around to watching the video version of it, what better time than now?
OpenAI says free and Plus subscribers can now use the feature without giving over their chats to train its models.
With chat history on, users can pick up previous chats where they left off, and the chatbot will reply as though they never stopped. The company also says users can start one-off chats that aren’t saved in the history.
At the moment, it seems Meta’s “AI studio” will let people make private and public bots, tuned for duties like personal shopping, trip-planning, meme generation, and helping users “never miss a romantic connection.” (I assume that last one is designed to trawl Craigslist Missed Connections for you.)
Alessandro Paluzzi posted these screenshots in a thread where he’s been tracking the feature since January.
Microsoft’s new AI model, VASA-1, transforms a single still image and an audio clip into an animated video, which is impressive, if not a little creepy.
The benefits – such as enhancing educational equity, improving accessibility for individuals with communication challenges, offering companionship or therapeutic support to those in need, among many others – underscore the importance of our research and other related explorations.
Microsoft says it won’t release a demo, API, or product with VASA-1 “until we are certain that the technology will be used responsibly.”
Meta’s AI assistant is being put everywhere across Instagram, WhatsApp, and Facebook. Meanwhile, the company’s next major AI model, Llama 3, has arrived.
OpenAI’s Batch API now lets users upload a file of bulk queries to the AI model, like categorizing data or tagging images, with the understanding that they won’t need immediate attention. Promising results within 24 hours lets them run when there is unused compute power, and keeps those pricey GPUs humming around the clock.
A bunch of places covered AI-generated images of an unholy Jesus/shrimp hybrid going viral on Facebook earlier this year, but the attention didn’t cause Zuck to take any action to slow the situation down. Here’s JC Noods laying back on a Lambo, a post which has 36,000 likes on Facebook right now. The AI internet is going great, y’all.
OpenAI chose Tokyo for its first office in Asia as it expands its footprint outside of the US. The company is also releasing a version of GPT-4 in Japanese.
Orifice.ai is where someone is tracking their progress in creating an apparently LLM-connected sex toy, complete with “generative moaning” and computer vision. (Spotted on Hacker News.)
NSFW, by the way.
[orifice.ai]
The Knight Rider Historians channel used ChatGPT, Amazon Polly, and a Raspberry Pi to make the world’s most annoying version of KITT, the AI that helps David Hasslehoff from its home inside the fictional Pontiac Firebird Trans Am from the Knight Rider series.
What do we think? 3 out of 10? 4?
ChatGPT Plus will be able to give more thought-out answers and understand longer prompts thanks to GPT-4 Turbo. The updated model, released on preview in November, trained on data until April 2023, so its knowledge base is also more recent.
Based on the video AssembleDebug shared below, it looks like it’ll work exactly as it does in iOS: tap the Gemini logo at the top, and you’ll be switched over to a chatbot prompt field. There, you can type to Google’s revamped chatbot, ask it to create images, or have it analyze pictures you send it.
The stock image company has had deals to let OpenAI, Meta, Google, and Amazon train on its images, Reuters wrote yesterday while reporting that Apple got in on the action.
Each of the deals reportedly initially ranged from $25 million to $50 million, but Shutterstock’s CEO, Jarrod Yahes, told the outlet most companies expanded them later.
A small group of users in the UK will start seeing AI-generated responses in Google Search, according to a report from the BBC. The AI answers will only appear for questions “where current trials have suggested they were helpful.”
Google started testing its Search Generative Experience (SGE) feature in the US last year, but it’s only available to those who opt-in to Search Labs.
