Tootfinder

A profile of Klay, whose app will let users remake songs using AI; Klay has raised ~$10M and is the first AI startup to close deals with the three major labels (Abram Brown/The Information)
https://www.theinformation.com/articles/secretive-startup-klay-jump…

Secretive Startup Klay Jumps Into the Fray Over AI Music
“So here’s a real surprise,” said Ary Attie, holding open a hidden door in the bowels of the Walker Hotel in the city’s Greenwich Village neighborhood. “There used to be a lot more cool, interesting places in New York—and there’s still a few left.” Attie, the 30-year-old founder of Klay, a new ...

A profile of Klay, whose app will let users remake songs using AI; Klay has raised ~$10M and is the first AI startup to close deals with the three major labels (Abram Brown/The Information)
https://www.theinformation.com/articles/secretive-startup-klay-jump…

Secretive Startup Klay Jumps Into the Fray Over AI Music
“So here’s a real surprise,” said Ary Attie, holding open a hidden door in the bowels of the Walker Hotel in the city’s Greenwich Village neighborhood. “There used to be a lot more cool, interesting places in New York—and there’s still a few left.” Attie, the 30-year-old founder of Klay, a new ...

Amazon Employees for Climate Justice strike again! https://www.wired.com/story/amazon-employees-open-letter-artificial-intelligence-layoffs/?emci=8691…

Amazon Workers Issue Warning About Company’s ‘All-Costs-Justified’ Approach to AI Development
Amazon Employees for Climate Justice says that over 1,000 workers have signed a petition raising “serious concerns” about the company’s “aggressive rollout” of artificial intelligence tools.

US-based AI startup Arcee releases Trinity Large, a 400B-parameter open-weight model that it says compares to Meta's Llama 4 Maverick 400B on some benchmarks (Julie Bort/TechCrunch)
https://techcrunch.com/2026/01/28/tiny

Tiny startup Arcee AI built a 400B-parameter open source LLM from scratch to best Meta's Llama | TechCrunch
30-person startup Arcee AI has released a 400B model called Trinity, which it says is one of the biggest open source foundation models from a U.S. company.

“[open social] must be funded like infrastructure, not consumer apps. In fact, given the scale of hybrid threats, it should be funded from defense budgets, not innovation leftovers.
The strategic layer of democracy is the information layer. And right now, it is outsourced.” @seabass.bsky.social
<…

Every business venture in history ever, even those with the most devastatingly high losses, was still more profitable than Open AI

From The Conversation
Canada’s long history with public service media offers a useful model for thinking about how AI could serve the public.
A publicly funded AI system could draw on public-domain materials, government datasets and openly licensed cultural content. It could be offered as an open-source system, making it widely available to researchers, developers and everyday users alike.

I spent some more time on AI coding evaluation tooling using the new #AWS open source Standard Operating Procedure system, and now I have an automated leader board. It rates the #Claude-flow hive mind version as better than the Claude-flow swarm version from a month before. Next step is to run some more …

With at least one powerful acceleratorist ai corporation looking to fund/partner with open source live coding projects, I wonder if anyone will take the bait

The Allen Institute for AI launches SERA, open-source coding agents including 32B- and 8B-parameter models designed to adapt to private codebases (Kyt Dotson/SiliconANGLE)
https://siliconangle.com/2026/01/27/ai2-launches-family-open-sou…

Ai2 launches family of open-source AI developer agents that adapt to any codebase - SiliconANGLE
Ai2 launches family of open-source AI developer agents that adapt to any codebase - SiliconANGLE

Essential AI is an interesting company, their Rnj-1 model illustrates another approach compared to the large AI companies.
#AI

OpenAI has hired 7 employees from AI coding startup Cline, including Nik Pash, who left Cline after posting "imagine the smell" on X regarding an xAI hackathon (Kevin McLaughlin/The Information)
https://www.theinformation.com/articles/op

OpenAI Hires at Least Seven Staffers From Coding Startup Cline
OpenAI has recently hired more than a half dozen employees from Cline, a startup that develops an open-source, AI-powered coding assistant, an OpenAI spokesperson confirmed. The hires could help OpenAI accelerate improvements for its Codex coding product, which it sees as a key tool to make its ...

🇺🇦 Auf radioeins läuft...
Air:
🎵 Sexy Boy
#NowPlaying #Air
https://djhoneyb.bandcamp.com/track/air-sexy-boy-honey-bs-emotional-garage-edit-2
https://open.spotify.com/track/6INLpBxo9F5QMer04VXEnd
Schon gewusst❓
Wenn Du einen Beitrag des Bots ⭐ FAVORISIERST, erfährt das nur der Bot - nicht Deine FollowerInnen.
Wenn Du einen Song weitererempfehlen möchtest, dann 🔁 TEILE ihn❗

Advice for all the poor clickworkers and AI agents who are trying to create accounts on half-open fediverse servers:

1. Do not begin your reasons for joining with the words, “I’m a genuine Fediverse user …”
2. Do not end the with, “… interested in technology, digital culture, and open communities. I’d like to join fedifreu.de to exchange ideas respectfully, follow diverse perspectives, and be part of a friendly, decentralized social space that values openness and co…

And finally, Chatgpt comes with ads!
Maybe try https://mistral.ai/ instead.
https://www.

Frontier AI LLMs, assistants, agents, services | Mistral AI
The most powerful AI platform for enterprises. Customize, fine-tune, and deploy AI assistants, autonomous agents, and multimodal AI with open models.

Taking up all the remaining oxygen in the room right now is AI hype,
exemplified by Jared Kaplan, who is quoted in this article:
-- ⭐️ --
"I spoke to Jared Kaplan, co-founder of Anthropic, the company behind the chatbot Claude.
He was a physicist the last time we spoke.
As a grad student at Harvard in the 2000s, he worked with the renowned theorist Nima Arkani-Hamed to open up the new directions in amplitude research that are being actively pursued today.

Is Particle Physics Dead, Dying, or Just Hard? | Quanta Magazine
Columnist Natalie Wolchover checks in with particle physicists more than a decade after the field entered a profound crisis.

DeepSeek launches DeepSeek-OCR 2, an upgraded optical character recognition model that replaces OpenAI-developed CLIP framework with Alibaba's Qwen2-0.5b (Ben Jiang/South China Morning Post)
https://www.scmp.com/tech/article/3341437/

DeepSeek taps Alibaba open-source AI technology to boost OCR performance
The Chinese AI start-up says its latest OCR model delivers stronger performance after adopting an Alibaba-developed open-source model.

I must say... :)
Meetily is the bees knees for transcribing online meetings where you cannot use the native transcription
https://meetily.ai/

“Clawdbot can execute Terminal commands, write scripts on the fly and execute them…”
I don’t even trust myself to do those things before 9am most days!
I just do not think I can be convinced I need a “digital assistant” in my life or on my computers (with elevated permissions, no less). https://mastodon…

Federico Viticci :ticciseal: (@viticci@macstories.net)
Attached: 1 image An open-source AI agent running on my Mac mini server is the most fun and productive experience I've had with AI in a while. Clawdbot by @steipete@mastodon.social is a personal AI assistant that you can infinitely tweak. It runs in any messaging app (!), has access to skills, can use the shell, and – notably – it can improve itself because it's completely open to customization. My story: Clawdbot Showed Me What the Future of Personal AI Assistants Looks Like https://ww…

I found this an interesting view on the AI scraping and Hoovering everything disease.
This view focuses on the overhead and costs that such hoovering places on our packet carrying infrastructure.
"The hiQ Decision Legalized Infrastructure Theft - We Need a Federal Fix"
https://

The hiQ Decision Legalized Infrastructure Theft - We Need a Federal Fix
The hiQ ruling erased legal protections against commercial scraping, leaving infrastructure providers to absorb escalating costs. Without federal action defining data misappropriation, a free-rider AI economy could undermine open networks, investment, and long-term data integrity.

🇺🇦 #NowPlaying on #BBC6Music's #LaurenLaverne
Air:
🎵 Sexy Boy
#Air
https://djhoneyb.bandcamp.com/track/air-sexy-boy-honey-bs-emotional-garage-edit-2
https://open.spotify.com/track/6INLpBxo9F5QMer04VXEnd

Humans are better coders than AI, CodeRabbit concludes
Security researchers analyzed 470 open-source GitHub pull requests, including 320 AI-co-authored PRs and 150 human-only PRs. In short, they found that AI tools accelerate code output, but also amplify the number of (critical) mistakes.
https://

Unit 42 details how underground hacking forums advertise and sell custom, jailbroken, and open-source AI hacking tools such as WormGPT and KawaiiGPT (Derek B. Johnson/CyberScoop)
https://cyberscoop.com/malicious-llm-tools-cybercrime-wormgpt-kawaiigpt/

"Public AI, Built On Open Source, Is The Way Forward In The EU". -Glyn Moody @…
This seems to opens the door to work on important problems, for instance building a solid climateGPT in Europe.

Libraries open their archives to train AI chatbots with books spanning centuries of human knowledge | Milwaukee Independent
https://www.milwaukeeindependent.com/newswire/libraries-open-archives-train-ai-chatbot…

Cash-flows in the #AIBubble, in one neat diagram. Carole Cadwalladr has one question, "when will it pop". No answer yet, but the question is delicious. https://open.substack.com/pub/broligar

#Magika 1.0: Smarter, Faster #FileDetection with Rust and AI
https://www.infoq.com/news/2025/12/magika-

Derrick Ross from Shakespeare reckons there open web is in trouble because you got to be a dev to build a website.
He seems to think his app Shakespeare can make it easier to make a web app by using ai 😕
"Build me a Twitter like website" is the kind of instruction he thinks it will handle.
A local app running on your own machine, though calling the big ai model apis. Including if you have the power at home to run deep seek or open models.
I find myself suspecting it'd be hard for vibe coders who aren't devs to tell if they had vibe coded buggy insecure software or not.
#vibeCoding #nostershire

Prime Intellect debuts INTELLECT-3, an RL-trained 106B parameter open source MOE model it claims outperforms larger models across math, code, science, reasoning (Prime Intellect)
https://www.primeintellect.ai/blog/intellect-3

INTELLECT-3: A 100B+ MoE trained with large-scale RL
Today, we release INTELLECT-3, a 100B+ parameter Mixture-of-Experts model trained on our RL stack, achieving state-of-the-art performance for its size across math, code, science and reasoning benchmarks, outperforming many larger frontier models.

Flock Exposed Its AI-Powered Cameras to the Internet. We Tracked Ourselves. https://www.404media.co/flock-exposed-its-ai-powered-cameras-to-the-internet-we-tracked-ourselves/

This website illustrates nicely how the US lost the competition–at least for now–in open(ish) LLM models: #AIResearch #AGI_hype
/via Wired

När AI kan läsa filer och samarbeta, inte bara chatta. Jag byggde Ogmios, en personlig AI-infrastruktur med specialiserade assistenter för kunskapsarbete. Om möjligheter, begränsningar och open source.
https://carlheath.se/fran-obsidian-til…

j’ai complètement open-sourcé tout le code qui fait tourner mon blog, évidemment je n’ai rien inventé (Š part les plugins) et j’en ai fait un tuto si d’aucun/es sont assez fou pour se lancer la dedans 😀 https://rmendes.net/content/articles/2026-01-24-deploy-your-own-…

Deploy Your Own IndieWeb Site on Cloudron with Indiekit
Deploy Your Own IndieWeb Site on Cloudron with Indiekit 24 January 2026 indiekit indieweb cloudron self-hosting Own your content. Syndicate everywhere. This guide walks you through deploying a complet...

The EU gives Google a six-month deadline to lift technical barriers to rival AI search assistants on Android and give key data to other search engine providers (Samuel Stolton/Bloomberg)
https://www.bloomberg.com/news/articles/2026-…

Google Android Told by EU to Open Up to Rival AI Systems
Google was handed a six-month European Union deadline to lift technical barriers to rival AI search assistants on Android and give key data to other search engine providers in the latest round of its Big Tech crackdown.

Sage advice about artificial intelligence: "Do your own original thinking first."
Because “Plans are worthless, but planning is everything.” 
https://open.substack.com/pub/philkoopman/p/plans-plan…

Plans, Planning, and Generative AI
The creative experience often matters more than the creative output. Be smart about how to use GenAI in that situation.

Extremely upset that a throw-away XKCD joke somehow became the organizing principle for the Internet.

AI is Dumber on Mondays
It's Monday morning. My coffee is hot, my IDE is open, and Claude Code just suggested I fix a bug by... creating the exact same bug in a different file. Cursor is hallucinating function names that don't exist. Codex thinks undefined is a valid return type in TypeScript.
Did the AI catch a case of the Mondays?
🤔

AI is Dumber on Mondays
It’s Monday morning. My coffee is hot, my IDE is open, and Claude Code just suggested I fix a bug by… creating the exact same bug in a different file. Cursor is hallucinating function names that don’t…

A look at Moltbook, a social network where OpenClaw assistants interact autonomously, as they discuss consciousness and identity, technical tips, and more (Simon Willison/Simon Willison's Weblog)
https://simonwillison.net/2026/Jan/30/moltbook/

Moltbook is the most interesting place on the internet right now
The hottest project in AI right now is Clawdbot, renamed to Moltbot, renamed to OpenClaw. It’s an open source implementation of the digital personal assistant pattern, built by Peter Steinberger …

We can at least be grateful to the AI bubble for helping open a vast number of eyes to the ensloppification of literally everything in all domains

"Although Ecosia is a not-for-profit, they seem to be re-enacting the for-profit erosion of human knowledge and the active murder of the open web currently being orchestrated by Google and the other tech giants."
(Original title: Ecosia’s “Green AI”: false hope for corrosive tech)
https://

Ecosia’s “Green AI”: false hope for corrosive tech
Why replace human work with fossil-fuel-slurping mistake generators? Ecosia’s “Green AI” is selling false hope for bad tech

"Artificial intelligence might be the most transformative technology in generations. It is also the most joyless.
While Wall Street greets AI with open arms, ordinary Americans respond with ambivalence, anxiety, even dread."
The Most Joyless Tech Revolution Ever: AI Is Making Us Rich and Unhappy

I particularly liked the summary "Understanding the Limits of Documentary Evidence". A #MustRead for #genealogy #familyhistory

What Happens When AI Analyzes a 160-Year-Old Murder Case—Without the Legend
For 160 years, the story of Tom Dula has been filtered through folklore, ballads, and legend.

Exciting news, friends: I’ll be joining an amazing lineup this week as MC for the Mozilla Festival in Barcelona! If you’ll be there in person, please do say hi, and if not, join online as the schedule promises to be extraordinary. Get the details here: https://www.mozillafestival.org/en/

Mozilla Festival 2025 | Ethical AI, Open Source & Digital Rights in Barcelona
Join MozFest 2025 in Barcelona, Nov 7–9 — the global AI policy summit, digital rights event & open source festival for a healthier internet.

Appfigures: Sora's app downloads dropped 32% month-over-month in December and 45% in January to 1.2M; consumer spending on the app fell 32% MoM as of January (Sarah Perez/TechCrunch)
https://techcrunch.com/2026/01/29/openais-sora-app-is-struggli…

OpenAI's Sora app is struggling after its stellar launch | TechCrunch
New data suggests the app is seeing declines in both app downloads and consumer spending, as the early hype about the AI video social network wears off.

In an open letter addressed to von der Leyen on Monday, more than 70 scientists call out the EU’s president for hyping AI, including two members of the UN’s high-level advisory body on AI
Stop overhyping AI, scientists tell von der Leyen | Euractiv

In UK parliamentary testimony, a Google executive says the company should not be required to license "freely available" open web content for AI training (Charlotte Tobitt/Press Gazette)
https://pressgazette.co.uk/platforms/google-ai-training-pay-opt-…

Google IP boss: We shouldn't pay for AI training on 'freely available' content
Google's global IP boss has said the tech giant does not believe it should have to pay to use "freely available content" for AI training.

I gave a lightning talk about the use of generative #AI at our engineering offsite this week. My general conclusion is that it is useful in some places but where you are a domain expert the gains might not be there. Just before the talk I came across this study:

Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity

"To be very clear, Gemini 3 isn’t perfect, and it still needs a manager who can guide and check it. But it suggests that “human in the loop” is evolving from “human who fixes AI mistakes” to “human who directs AI work.” And that may be the biggest change since the release of ChatGPT."

Three Years from GPT-3 to Gemini 3
From chatbots to agents

Character.AI launches Stories, which lets users create interactive fiction and is the company's "first new format created specifically" for users under 18 (Amanda Silberling/TechCrunch)
https://techcrunch.com/2025/11/25/char

Character AI will offer interactive 'Stories' to kids instead of open-ended chat | TechCrunch
The company announced last month that it would no longer allow minors to use its chat features.

Introducing #Mistral 3
#LLM

From various rumors and sources it appears that Open AI - a company organized under Delaware law but with HQ in California - is the cause of the tripling of computer memory prices and shortages.
Word is that Open AI bought up all of the large wafers from Samsung and Hynix that are used to make memory. The intention seems to be to corner the market in order to hobble Open AI's competition.
I am not an expert in either Federal or State laws on anti-competitive practices.

Chinese AI startup Z.ai releases GLM-4.7, an open-weight model that Z.ai says delivers significant improvements in coding performance compared to GLM-4.6 (Z.ai)
https://z.ai/blog/glm-4.7

Hoe China de wereld van open AI modellen domineert, mooi overzicht. Zonder dat je het weet zit achter de toepassingen van veel bedrijven een chinees open model, al dan niet verbeterd/gefinetuned en dergelijke. https://www.interconnects.ai/p/8-plots-that-explain-the-state-of…

8 plots that explain the state of open models
Measuring the impact of Qwen, DeepSeek, Llama, GPT-OSS, Nemotron, and all of the new entrants to the ecosystem.

huh, based

Over the last decade, America’s roads have become more dangerous,
with serious crashes increasing by nearly 20 percent since 2013.
Approximately 94 percent of crashes are the result of driver behavior
like speeding, impairment or distraction
— behavior that can be detected and corrected by a new generation of machine learning-enabled dash-cams.
Seamless integration between machine learning, IoT management and the cloud allows these cameras to improve safety in r…

Oracle stock fell 30% this quarter, its steepest drop since Q3 2001, when it slid ~34%, amid skepticism about its ability to open more data centers for OpenAI (Jordan Novet/CNBC)
https://www.cnbc.com/2025/12/26/oracle-stock-on-pace-for-worst…

Oracle shares on pace for worst quarter since 2001 as new CEOs face concerns about AI build-out
Investors want to know if Oracle, under new CEOs Clay Magouyrk and Mike Sicilia, can pay for and deliver data centers packed with Nvidia chips for OpenAI.

A look at Clawdbot, an open-source personal AI agent that runs locally on the user's computer and integrates with multiple LLMs and messaging services (Federico Viticci/MacStories)
https://www.macstories.net/stories/clawdbot-sho…

Clawdbot Showed Me What the Future of Personal AI Assistants Looks Like
For the past week or so, I’ve been working with a digital assistant that knows my name, my preferences for my morning routine, how I like to use Notion and Todoist, but which also knows how to control Spotify and my Sonos speaker, my Philips Hue lights, as well as my Gmail. It runs on...

Q&A with Z.ai Director of Product Zixuan Li on Chinese AI models embracing open source, attracting global users for its GLM model, training on memes, and more (ChinaTalk)
https://www.chinatalk.media/p/the-zai-playbook

The Z.ai Playbook
Zixuan Li on the Chinese AI ecosystem

OpenAI details efforts to secure its ChatGPT Atlas browser against prompt injection attacks, including building an "LLM-based automated attacker" (Rebecca Bellan/TechCrunch)
https://techcrunch.com/2025/12/22/openai-say…

OpenAI says AI browsers may always be vulnerable to prompt injection attacks | TechCrunch
OpenAI says prompt injections will always be a risk for AI browsers with agentic capabilities, like Atlas. But the firm is beefing up its cybersecurity with an "LLM-based automated attacker."

Reading thoughts about a new Chinese openweights AI model, Kimi K2 Thinking.
https://www.interconnects.ai/p/kimi-k2-thinking-what-it-means

5 Thoughts on Kimi K2 Thinking
Quick thoughts on another fantastic open model from a rapidly rising Chinese lab.

Lightning AI merges with data center operator Voltage Park to create an "AI cloud" with a $2.5B valuation, managing 35,000 Nvidia GPUs across six data centers (Iain Martin/Forbes)
https://www.forbes.com/sites/iainmartin/20

AI Startup Merges With A Billionaire-Backed Data Center Operator In $2.5 Billion Deal
The startup behind open source tool PyTorch Lightning has merged with compute provider Voltage Park to create a “full stack AI cloud” to serve corporates and startups like Cursor.

Analysis: since 2023, data center power demands have delayed 15 coal plants' retirements; the Trump administration has ordered two power plants to remain open (Ariel Wittenberg/Politico)
https://www.politico.com/news/2025/11/27/a

Google DeepMind plans to open an AI research lab in Singapore, collaborating with institutions across the Asia-Pacific in education, public services, and more (Lila Ibrahim/Google DeepMind)
https://deepmind.google/blog/were-expa…

Google DeepMind opens Singapore research lab for Asia-Pacific AI.
Google DeepMind opens a new research lab in Singapore to accelerate the development of frontier AI across the Asia-Pacific region through research, talent, and strategic collaborations.

Essential AI, whose CEO co-wrote Google's Attention Is All You Need paper, unveils Rnj-1, an 8B-parameter open model with SWE-bench performance close to GPT-4o (Ashish Vaswani/Essential AI)
https://essential.ai/research/rnj-1

Chinese startup Moonshot updates its Kimi model, saying K2.5 can process text, images, and videos simultaneously and beats its open-source peers in some tests (Bloomberg)
https://www.bloomberg.com/news/articles/2026-01-27/chi…

Boltz, which is developing open frontier AI models for biomolecular research, raised a $28M seed led by Zetta, Amplify, and a16z at a valuation of $125M (Andrew Dunn/Endpoints News)
https://endpoints.news/exclusive-boltz-raises-28m-seed-to-build-ope…

Boltz raises $28M seed to build open AI models in biology
Boltz raises $28M seed round to develop open-source AI models for drug discovery, partners with Pfizer. Founded by MIT researchers after AlphaFold 3's closed release.

A look at Confer, an open-source AI assistant project from Signal creator Moxie Marlinspike that is designed to provide end-to-end encryption for AI chats (Dan Goodin/Ars Technica)
https://arstechnica.com/security/2026/0…

Signal creator Moxie Marlinspike wants to do for AI what he did for messaging
Introducing Confer, an end-to-end AI assistant that just works.

MIT and Hugging Face study: the share of downloads of Chinese-made open AI models rose to 17% in the past year, surpassing US developers' 15.8% share, a first (Melissa Heikkilä/Financial Times)
https://on.ft.com/4ok6Sqb

Allen Institute for AI, or Ai2, unveils Olmo 3 models that it says outperform open models like Stanford's Marin and commercial open-weight models like Llama 3.1 (Todd Bishop/GeekWire)
https://www.geekwire.com/2025/ai2-releases

Google releases TranslateGemma, a suite of Gemma 3-based open translation models available in 4B, 12B, and 27B parameter sizes with support for 55 languages (The Keyword)
https://blog.google/innovation-and-ai/technology/developers-tools/translategemma/

TranslateGemma: A new suite of open translation models
TranslateGemma is a new family of open translation models built on Gemma 3.

Zoom says its "federated AI" model, combining its SLM with open- and closed-source models, got 48.1% on Humanity's Last Exam vs. 45.8% for Gemini 3 Pro w/ tools (Xuedong Huang/Zoom)
https://www.zoom.com/en/blog/humanitys-last-exam-zoom-ai-breakthrough/

Zoom AI sets new state-of-the-art benchmark on Humanity's Last Exam
Federated innovation driving breakthrough results in complex AI testing

Nvidia announces it has acquired SchedMD, the developer of Slurm, an open-source workload management system for HPC and AI (Arsheeya Bajwa/Reuters)
https://www.reuters.com/business/nvidia-buys-ai-software-provider-schedmd-expan…

Mistral's CEO says claims that Chinese AI startups lag behind the US are a "fairy tale," and China's open-source tech is "probably stressing" US CEOs (Bloomberg)
https://www.bloomberg.com/news/articles/2026-0…

Chinese AI startup Z.ai releases the GLM-4.6V series of vision models, with support for native function calling, available with 106B and 9B parameters (Carl Franzen/VentureBeat)
https://venturebeat.com/ai/z-ai-debuts-open-source-glm-4-6v-a-n…

An analysis of 100T tokens from the past year shows reasoning models now represent over half of all usage, open-weight model use has grown steadily, and more (OpenRouter)
https://openrouter.ai/state-of-ai

State of AI | OpenRouter
An empirical study analyzing over 100 trillion tokens of real-world LLM interactions across tasks, geographies, and time.

Runway launches Gen 4.5, a new text-to-video AI model that produces HD videos from written prompts and excels at physics; Gen 4.5 tops Video Arena's leaderboard (Ashley Capoot/CNBC)
https://www.cnbc.com/2025/12/01/runway-gen-4-5-video-model-google-open-ai.h…

Runway rolls out new AI video model that beats Google, OpenAI in key benchmark
Gen 4.5, a new AI model that allows users to generate high-definition videos based on written prompts.

In South Korea's competition to develop an indigenous AI model, three of the five finalists used foreign open-source code, which they argue is practical (Jiyoung Sohn/Wall Street Journal)
https://www.wsj.com/tech/ai/…

Meta introduces Omnilingual Automatic Speech Recognition, a suite of AI models providing automatic speech recognition capabilities for more than 1,600 languages (Carl Franzen/VentureBeat)
https://venturebeat.com/ai/meta-returns-to-open-source-ai…

Nvidia announces Alpamayo-R1, an AI model for autonomous driving research, and calls it the "first industry-scale open reasoning vision language action model" (Rebecca Szkutak/TechCrunch)
https://techcrunch.com/2025/12/01/nvid

Nvidia announces new open AI models and tools for autonomous driving research | TechCrunch
Nvidia continues its push into physical AI with the release of a new reasoning world model and other tools for physical AI.

Anthropic makes Claude Cowork available to $20/month Pro subscribers, after launching it for Max users, and says Pro users may hit their usage limits sooner (Ian Carlos Campbell/Engadget)
https://www.engadget.com/ai/anthropic-open

Anthropic opens up its Claude Cowork feature to anyone with a $20 subscription
Claude Cowork is now available to anyone subscribed to Anthropic's Pro plan.

The Allen Institute of AI launches Bolmo 7B and Bolmo 1B, claiming they are "the first fully open byte-level language models", built on its Olmo 3 models (Emilia David/VentureBeat)
https://venturebeat.com/ai/bolmos-architecture-unlock…

As part of the AWS deal, OpenAI says it will immediately begin running workloads on AWS infrastructure, tapping hundreds of thousands of Nvidia GPUs in the US (MacKenzie Sigalos/CNBC)
https://www.cnbc.com/2025/11/03/open-ai-amazon-aws-cloud-deal.html

Amazon closes at record after $38 billion OpenAI deal with AWS
OpenAI will immediately begin tapping Nvidia's AI processors using infrastructure from Amazon Web Services.

Chinese startup Moonshot releases Kimi K2 Thinking, an open-source model it claims beats GPT-5 in agentic capabilities; source: the model cost $4.6M to train (Evelyn Cheng/CNBC)
https://www.cnbc.com/2025/11/06/alibaba-backed-moonshot-releas…

Alibaba-backed Moonshot releases its second AI update in four months as China’s AI race heats up
The Chinese AI startup on Thursday released its newest "Kimi K2 Thinking" artificial intelligence model.

Z.ai releases GLM-Image, an open-source multimodal AI model trained on Huawei chips that it says is China's first to be fully trained using domestic chips (Luz Ding/Bloomberg)
https://www.bloomberg.com/news/articles/2026-01-…

The EU expects to launch a formal bidding process for its AI gigafactories in early 2026 and close it in summer 2026, as the bloc seeks to catch up with the US (Mauro Orru/Wall Street Journal)
https://www.wsj.com/tech/ai/eu-to-open-…

The current AI strategies of China and the US are complementary, as unlike the US, China isn't "AGI-pilled" yet, focusing on embodied AI and open source models (Dean W. Ball/Hyperdimensional)
https://www.hyperdimensional.co/p/the-bitter-lessons

The Bitter Lessons
Thoughts on US-China Competition

An investor group and AI developer Voltai plan to use AI to design, build, and run a 3GW data center in South Korea, set to cost up to $35B and open in 2028 (Jiyoung Sohn/Wall Street Journal)
https://www.wsj.com/tech/a-…

US startups are increasingly adopting Chinese open source AI models, as some capabilities of Chinese models start to catch up with those of frontier US labs (NBC News)
https://www.nbcnews.com/tech/innovation/silicon-valley-building-free-chinese-ai-r…

More of Silicon Valley is building on free Chinese AI
Surveying the state of America’s artificial intelligence landscape earlier this year, Misha Laskin was concerned

Starcloud, which launched a satellite with a Nvidia H100 chip in November, says the satellite is running and querying responses from Google's Gemma (Pia Singh/CNBC)
https://www.cnbc.com/2025/12/10/nvidia-backed-starcloud-tr…

‘Greetings, earthlings’: Nvidia-backed Starcloud trains first AI model in space as orbital data center race heats up
The company's Starcloud-1 satellite is running Gemma, an open model from Google, marking the first time in history that an LLM has been trained in outer space.

Source: Anthropic, OpenAI, Google, Microsoft, and more are set to unveil the Agentic AI Foundation to build open-source AI agent standards as soon as this week (Aaron Holmes/The Information)
https://www.theinformation.com/articles/openai…

OpenAI, Anthropic, Google Agree to Develop Agent Standards Together
For AI agents to work properly in automating white-collar tasks, the companies developing the agents and the companies running the enterprise apps those agents use will need to agree on technical standards for how these technologies connect to each other.Some leading companies are preparing to ...

How Google's ambitious approach to training Gemini on text, code, audio, images, and video helped it stage a powerful comeback, triggering a Code Red at OpenAI (Katherine Blunt/Wall Street Journal)
https://www.wsj.com/tech/ai/google-ai-open

Anthropic open sources a method to score AI model political evenhandedness; Gemini 2.5 Pro got 97%, Grok 4 96%, Claude Opus 4.1 95%, GPT-5 89%, and Llama 4 66% (Ina Fried/Axios)
https://www.axios.com/2025/11/13/anthropic-bot-bias-data

Anthropic tries to measure bot bias
There's no consensus over what constitutes political bias in AI systems.

AWS launches Nova Forge, a $100,000/year service allowing clients to customize Amazon's AI models at various stages of training and refine open-weight models (Jordan Novet/CNBC)
https://www.cnbc.com/2025/12/02/amazon-nova-forge-le…

Amazon to let cloud clients customize AI models midway through training for $100,000 a year
It can cost hundreds of millions or even billions to build a large language model from scratch, and this process is more affordable, Amazon said.

OpenAI says ChatGPT will now ditch em dashes if users tell it to; em dashes have become telltale signs that supposedly signals text written by AI (Sarah Perez/TechCrunch)
https://techcrunch.com/2025/11/14/openai-says-its-fixed-chatgpts-em-dash-problem/

OpenAI says it's fixed ChatGPT's em dash problem | TechCrunch
OpenAI says users can now personalize ChatGPT to stop using the em dash in its output.

Sources: Meta's new AI model, codenamed Avocado, may launch in spring 2026 as a "closed" model, and was trained using Google's Gemma, OpenAI's gpt-oss, and Qwen (Bloomberg)
https://www.bloomberg.com/news/articles/20

Unsealed docs from Elon Musk's OpenAI lawsuit, set for a jury trial on April 27, show Sutskever's concerns about treating open-source AI as a "side show", more (Alex Heath/The Verge)
https://www.theverge.com/column/863319/highlights-musk-v-altman-open…

‘Sideshow’ concerns and billionaire dreams: What I learned from Elon Musk’s lawsuit against OpenAI
Highlights from Musk v. Altman.

Sources: Meta's new model, codenamed Avocado, may launch in 2026 as a "closed" model, a move Alexandr Wang advocates, shifting from Meta's open-source strategy (Bloomberg)
https://www.bloomberg.com/news/articles/20

The future of AI development could resemble the rivalry between iOS and Android, with China's AI approach closer to Google's open and customizable Android OS (Kai-Fu Lee/Financial Times)
https://www.ft.com/content/b1f92b0e-d6ef-4c95-b51e-7bcf90c8a65f
<…