Wes Roth
@WesRothMoney
FOLLOWS YOU. Artificial Intelligence, Automation & Optimism. Everything I say is 100% serious...
Qwen just released the Open Source BEAST 🔥🔥 Qwen3-Coder, their most agentic code model to date. beats KIMI K2 (!) multiple models, the most powerful is: Qwen3-Coder-480B-A35B-Instruct -256K tokens, extendable up to 1M tokens -Significant Performance among open models on…
Amazon is stepping up its AI game with the acquisition of Bee, the startup behind a $50 AI-powered wearable that listens, summarizes, and organizes daily tasks. With Alexa now upgraded with generative AI and Bee joining the family, Amazon is clearly aiming to turn personal…
OpenAI and Oracle are doubling down on AI infrastructure with a new 4.5 GW expansion of their massive Stargate project. This pushes Stargate to over 5 GW of data center capacity in development powered by 2+ million chips all aimed at keeping the U.S. ahead in the AI arms race.…
Imagine the combined brainpower of those AI giants. Rivals today, but allies tomorrow? 🤔
what if I told you that OpenAI, Google, Anthropic and xAI will all be working together in a few years
Yuval Noah Harari argues that compassionate AI cannot emerge from competition. If AI development is driven by power-hungry rivals in an arms race, AI will inherit these traits. Telling AI to be benevolent is meaningless if it observes humans lying, cheating, and competing…
Pro users in the EEA and Switzerland already have full access to ChatGPT’s Agent Mode, and Plus users worldwide are now starting to see the rollout. Who else is ready to see what these agents can really do?
Pro users in the European Economic Area and Switzerland, thanks for your patience — ChatGPT agent is now fully rolled out to you. And the rollout to Plus users globally has started and will continue over the next few days.
A new research paper from Anthropic and collaborators has revealed a surprising twist in AI reasoning: more compute at test time can actually make large reasoning models worse. The study shows that extending a model's reasoning chain often leads to distractions, overfitting, or…
With its Comet AI browser, Perplexity is talking to smartphone manufacturers to pre-install Comet and challenge Chrome and Safari’s grip on mobile devices. browser on new devices. With 2M daily users, a $14B valuation, and backing from Jeff Bezos and Eric Schmidt, Perplexity is…
Microsoft has poached over 20 DeepMind experts, including Amar Subramanya, the former engineering head of Gemini. With Mustafa Suleyman leading the charge, the company now boasts AI veterans like Amar Subramanya and Adam Sadovsky, signaling an aggressive push to rival Google’s…
Meta has just hired three more researchers from Google DeepMind — Tianhe Yu, Cosmo Du, and Weiyue Wang — all of whom worked on the Gemini model that hit gold-medal performance at the International Math Olympiad. With six DeepMind experts now onboard, Meta is clearly doubling…
Max Winga warns that superintelligence could threaten human existence if not carefully controlled. Everyone must work together globally to prevent this outcome. The specific approach—whether a UN convention or international treaty—matters less than achieving collective action.…
Eric Schmidt explains that general intelligence—AI with free will and strategic thinking—could emerge in 4–6 years. The next step, superintelligence, will surpass all human intelligence combined. A true test of superintelligence would be an AI proving something humans know is…
According to WSJ, Elon Musk’s xAI is teaming up with Valor Equity Partners to raise another $12 billion for Colossus 2, just weeks after raising $10B With massive data center plans and a compute target of 50M H100-equivalent units, Musk is clearly betting big on infrastructure…
CEO Dario Amodei confirmed the company will now pursue investments from Gulf states like the UAE and Qatar—despite previously rejecting such funding. He admitted it’s “not ideal” but necessary to keep pace in the AI arms race.
should all data centers and all employees working on AI be under 24/7 surveillance? this is what I mean when people ask me my p(doom) and I start talking about p(1984)
MIRI had just published the most Orwellian proposal possible: · Government surveillance of datacenters via hardware-enabled mechanisms and software-based tools. · Tracking personnel: Surveillance of key AI researchers and their locations, computers, and research activities. WTF
xAI is aiming for 50 million H100-equivalent units within 5 years, while OpenAI plans to cross 1 million GPUs online by year’s end.
The @xAI goal is 50 million in units of H100 equivalent-AI compute (but much better power-efficiency) online within 5 years