Analytics DeepLearning.AI - The Batch | DeepLearning.AI (Page 17)

A demonstration of video editing through text input, altering a runner’s background and costume.

Machine Learning Research

Familiar Faces, Synthetic Soundtracks: Meta debuts Movie Gen for text-to-video generation with consistent characters

Meta upped the ante for text-to-video generation with new systems that produce consistent characters and matching soundtracks.

Three people, Jeff Dean, Andrew Ng, and Geoff Hinton, stand together at Hinton’s retirement party.

Letters

Nobel Prizes for AI!: Nobel committees for physics and chemistry honored AI researchers. How can the AI community itself can celebrate the next generation of innovators?

Congratulations to Geoff Hinton and John Hopfield for winning the 2024 Physics Nobel Prize!

The Batch Newsletter

How Meta’s Movie Gen Does It, AI’s Criminal Underground, Court Says LAION is Legal, OpenAI’s New Voice API

The Batch AI News and Insights: Congratulations to Geoff Hinton and John Hopfield for winning the 2024 Physics Nobel Prize!

A modern building entrance where a human receptionist and a humanoid robot assist incoming visitors.

Data Points

Apple’s new models punch above their weight: Liquid explores a new multimodal modal architecture

Copilot may lead to more bugs than productivity gains. OpenAI prunes its Whisper model for faster completions. A new study measures top AI companies’ redteaming efforts. An open-source CLI coding assistant, o1-engineer.

A serene outdoor scene where a person is painting on a large canvas set against a beautiful natural landscape.

Data Points

Emu3 claims “next-token prediction is all you need”: Black Forest updates FLUX image model, adds a developer API

OpenAI introduces new canvas UI for editing with ChatGPT. Google’s chip design model gets a new (but familiar) name. Microsoft launches some of the features it promised with Recall. Aider wants to give the right models the right jobs.

California Senate Bill 1047 with a large veto stamp on top of the document.

Letters

A Victory for Innovation and Open Source: California’s governor vetoed SB 1047, which would have hurt innovation without making anyone safer. Now let’s work on mitigating AI’s real risks and realizing its potential benefits.

We won! California’s anti-innovation bill SB 1047 was vetoed by Governor Newsom over the weekend.

The Batch Newsletter

Llama Goes Multimodal, Pros Embrace Generative Video, Military AI Guidelines, LLMs That Read Spreadsheets

The Batch AI News and Insights: We won! California’s anti-innovation bill SB 1047 was vetoed by Governor Newsom over the weekend.

A GIF showcasing a dynamic spreadsheet interaction using AI, with cells being populated and analyzed automatically.

Machine Learning Research

Enabling LLMs to Read Spreadsheets: A method to process large spreadsheets for accurate question answering

Large language models can process small spreadsheets, but very large spreadsheets often exceed their limits for input length. Researchers devised a method that processes large spreadsheets so LLMs can answer questions about them.

Global leaders discuss Responsible AI in the military at the REAIM 2023 summit in The Hague, Netherlands.

Tech & Society

International Guidelines for Military AI: Global coalition endorses blueprint for AI’s military use

Dozens of countries endorsed a “blueprint for action” designed to guide the use of artificial intelligence in military applications.

A dynamic GIF featuring erupting volcanoes, a reindeer in the snow, animated fuzzy creatures, and a close-up of a human eye.

Tech & Society

Generative Video in the Editing Suite: Adobe integrates AI video generation into Premiere Pro

Adobe is putting a video generator directly into its popular video editing application.

An interior design assistant tool analyzing an image of a modern living room.

Machine Learning Research

Llama Herd Expands: Meta updates Llama models with vision-language, edge sizes, and agentic APIs

Meta extended its Llama family of models into two new categories: vision-language and sizes that are small enough to fit in edge devices.

A modern recording studio with multiple screens, displaying advanced AI transcription tools.

Data Points

Nemotron models boost Llama’s speed but maintain accuracy: NotebookLM “reads” audio and video

Hacking ChatGPT’s long-term memory function. U.S. trade commission targets companies who lie about AI. A new OpenAI model for screening text and images. Apple, Meta, and others hold off on AI Pact