November, apart from winds and Thanksgiving turkey, has brought many AI-related updates to popular tools used by companies around the world. This includes the Windows Notepad, but not alone.
This edition also delivers information about bigger models from China and the uneasy relationship between publishing companies and AI, including both collaboration and court battles.
4 Nov – Dutch publisher announced it will use AI translator
Veen Bosch & Keuning (VBK), the biggest publisher in the Netherlands, has announced that it will use AI-powered translations in their published books. The text project will be run on 10 books, all commercial fiction, with no works of literature being translated this way.
More can be found in The Guardian.
6 Nov – Microsoft enriches Notepad with generative AI
The simplest of the simplest text editing software – the notepad, around since 1983, got enhanced with generative AI. The AI-powered feature is called “rewrite” and, as the name states, it rewrites content delivered by the user by changing sentences, tone, or formatting. It can, for example, make a text longer or shorter, depending on the needs.
More can be found in TheVerge.
8 Nov – Google rolls out Vids AI
Vids AI is Google’s new app that enables users to create video presentations using a prompt. The software will automatically assemble a video presentation using available assets, including stock photos, images and videos uploaded by the user, as well as text. Vids will be available for Workspace organizations.
More can be found in TheVerge.
11 Nov – Google’s DeepMind Outsources AlphaFold 3
The model that solved the protein bending mystery has been open-sourced by Google’s DeepMind lab. The model was the reason why DeepMind researchers Demis Hassabis and John Jumper were awarded with 2024 Nobel Prize in chemistry.
More can be found in VentureBeat.
12 Nov – YouTube tests music remixes done with AI
A new feature that was recently rolled out for testing enables selected youtube creators to remix songs of their choice using AI. The remixes may be used only in YouTube Shorts – videos that are no longer than 30 seconds. The AI system may change a song’s speed, mood, or genre.
More can be found in TheVerge.
13 Nov – Wall Street Journal is testing AI text summaries
The renowned business newspaper is testing a “key points” box above texts published in the digital format. The box also has clear information that “Artificial Intelligence has created the summary.”
More can be found in TheVerge.
14 Nov – NASA launches AI Earth Copilot
Together with Microsoft, NASA has created an AI chatbot designed to provide the user with easily-digestible answers. The users may ask about nearly anything about Earth that is related with NASA data – for example, about clouds, oceans, or weather.
More can be found in the official announcement.
15 Nov – Google lets users prepare clipart with AI
Google is infusing its productivity suite with the GenAI model Gemini. With the new addition, the company allows users to create clipart to enrich documents they are working on with just a prompt. The feature is available through Google Workspace paid plans.
More can be found in TheVerge.
19 Nov – Bloomberg: Microsoft licenses HarperCollins Books
HarperCollins, one of the world’s leading publishers, has announced a three-year deal that allows authors to opt-in for a program where they get paid for submitting their works for training AI models. According to Bloomberg news, the unnamed company that struck a deal with HarperCollins is Microsoft, which aims to develop a yet-unnamed model using this data.
More can be found in Bloomberg.
22 Nov – Chinese researchers unveil LLaVA-o1
Researchers from several Chinese universities codeveloped LLaVA-o1, a new vision model that aims to use more advanced techniques (like Chain-of-Thought). Instead of immediate generation, the model aims to start reasoning and produce improvements during the process.
More can be found in VentureBeat.
27 Nov – Google launches AI-powered chess with customizable pieces
Google Labs has launched an online chess service that uses generative AI to deliver fully customizable pieces. The chat-based interface lets the user provide the system with some ideas on the looks and the system delivers the final design. The pieces can be inspired by crystals, mushrooms, trees, Roman architecture, or basically anything. The user may customize both parties, so it can be a fire vs. water match, skeletons vs. mushrooms, or anything else.
The new game is available on the Google Labs website.
29 Nov – Canadian news company sues OpenAI
News Media Canada, one of the largest Canadian news organizations, sues OpenAI for using their texts for training AI models without permission. According to a News Media Canada statement, OpenAI “cannibalizes proprietary content” and takes a free-ride on the back of publishers who “invest real money and employ real journalists who produce real stories for real people.”
More can be found in The Guardian.
29 Nov – Alibaba releases Qwen with Questions
Qwen with Questions is a new model from the growing Qwen family developed by Chinese e-commerce giant, Alibaba. According to recent benchmarks, the model is a good competitor for OpenAI’s GPT o1 model, mostly due to improved reasoning skills. By that, the model is better suited to tasks that require logical or sequential thinking, like coding or math.
More can be found in VentureBeat.
