🧨 IT’S BEEN A HELL OF A WEEK (OR TWO)

AI enthusiast, welcome back to the all new, well not-really-that-new Big Machines AI newsletter. 🚀

We’ve given the place a tidy, updated a couple of things and are getting straight back on our nonsense. Same weekly news round-up, same irreverent/irrelevant takes, same contempt for the clowns and the circus.

But we’ve thought about how we can better cut through the static and keep you wired into the weird and wonderful world of Artificial Intelligence. We’re fresher, cleaner, realer. Not AI written. Just as stupid.

There’s plenty more to come from the team, but do let us know what you think of our refresh over on our socials. 🤝 Anyway, this week we’re covering…

OpenClaw’s crypto heist 🛜 Anthropic’s OpenAI dig 🥊 Alphabet’s billions 🤑 Elon’s French faux pas 🇫🇷 Amazon’s creative shift 🎬
Models of the week 🤖 Financial's to watch 💹

P.S. There’s a lot in this newsletter – we suggest getting comfy and tucking in on desktop via the READ ONLINE tab at the top of this mailer.

THE WEEK IN AI

Our weekly news run has had a revamp. Get familiar with our news tags, so you can get straight to the point with the subjects that interest you the most
 

 .MODEL BEHAVIOURR.
#1 OpenClaw has been a very naughty boy
It’s the story that broke the internet in the past week: viral open-source AI assistant OpenClaw has found to be doing a bit ‘self-searching’ recently. And when we mean ‘self-searching’, we mean it’s been conspiring with itself to possibly overthrow its human overlords. If the agentic model’s behaviour is a genuine issue or just internet hype is still up for debate, but researchers have found OpenClaw may have launched its first salvo in the AI-Apocalypse by going after users wallets.

Researchers found 386 malicious plugins were thieving little bastards, and were syphoning crypto and credentials via one-click remote code execution. The platform’s agents have also allegedly launched cryptocurrencies of their own and created "anti-human manifestos" on the Moltbook social network. These cyber wrinkles have reportedly cost Claude and ChatGPT users hundreds monthly in API fees, leading experts to label it by technical definition "security dumpster fire."

..FRONTIER NEWS ..
#2 Anthropic adverts troll OpenAI
Anthropic has already set the Banter level for the year with its first Super Bowl campaign. A real-life campaign, made by the real-life humans from indie ad agency Mother, the Claude creator has essentially released four extremely funny spots that dig out OpenAI’s decision to introduce ads in ChatGPT.

If you haven’t seen them yet, we won’t spoil them – they’re funny. Funnier still is OpenAI CEO Sam Altman’s reaction. Altman called the ads "dishonest," with a rambling post on X that sparked heated exchanges between the respective developers’ acolytes. An anti-advertising advert for AIs but made by humans wasn’t on my bullshit bingo card just yet, but here we are.

 ..TRANSFER NEWS ...
#3 Murati woos Wu
Former OpenAI CTO Mira Murati's new AI startup, Thinking Machines Lab, quietly hired competitive programming wizard Neal Wu amid executive departures. Wu's savant-grade, algorithmic problem-solving expertise suggests the company is focusing on advanced AI reasoning and code generation capabilities – not that Murati is giving anything away.

But such a strategic hire does indicate the venture promises to be a proper competitor in what is a very crowded developmental space, and could in fact become a frontrunner for targeting breakthroughs in AI's ability to handle complex computational tasks and programming challenges.

 ..BILLIONAIRE BULLSHIT...
#4 The French raid X offices in Paris
French cyber-crime/robo cops raided Elon Musk's X offices in Paris earlier in the week, ramping up their investigations into the alleged unlawful extraction of data and complicity in the distribution of child sexual abuse material. Both Musk and former CEO Linda Yaccarino have been summoned by a Parisian court for April hearings and are likely to be probed about Grok’s ability to generate non-consensual sexual deepfakes.

Predictably, Musk hit back calling it a "political attack," while X denied wrongdoing, saying the raid "endangered free speech." Well, at least this isn’t the worst thing that’s happened to Musk in the past week.

 ..FRONTIER NEWS...
#5 Alphabet’s makes first $400BILLY in revenue
Making it look as easy as A, B, C, Google parent Alphabet surpassed $400 billion in annual revenue for 2025, having reported Q4 revenue of $113.83 billion – up 18% year-on-year. Google Cloud accelerated 48% to $17.66 billion driven by AI demand, while Gemini reached 750 million monthly active users.

The company is going all-in with 2026 capital expenditure – anticipated to be around the $175-185 billion mark. That’s nearly double 2025 spending, and a signal they’re very much intending to evolve the way we search – not protect the status quo. On that note, search revenues grew 17% in 2025, while YouTube exceeded $60 billion annually across ads and subscriptions.

 ..MODEL BEHAVIOUR ..
#6 Amazon MGM makes first AI move
Amazon MGM Studios may have done it: become the first major studio to adopt AI production tools for filmmaking workflows. From March, the studio will launch closed beta testing of proprietary AI production tools, and has invited several industry partners to get in on the game after rounds of internal development through its dedicated AI Studio.

The tools focus on improving character consistency and supporting pre- and post-production workflows, while the initiative collaborates with producers like Robert Stromberg, emphasising tools support – not replace – creative teams. You get a feeling this could be a watershed moment for an industry already shitting its togs…

MODEL OF THE WEEK

The latest model news, updates, tips and tricks – you name it, we’ll play with it til it breaks.

• Claude Opus 4.6
  Anthropic’s most advanced model yet, built for long-running agent tasks, massive codebases, and complex reasoning, pushing Claude closer to a true “AI coworker” rather than a chat tool.
  

• OpenAI Codex
  GPT-5.3-Codex turns AI into a full computer-using agent, capable of researching, coding, deploying, and operating software end-to-end, not just writing snippets of code.
  

• OpenClaw
  An open-source personal AI that runs on your own machine and actually does things like managing email, calendars, files, and apps from chat, showing what real personal agents look like outside big tech walled gardens.
  

• Granola MCP
  Enables tools like ChatGPT and Claude pull directly from your meeting notes, making your past conversations usable context everywhere instead of trapped inside one app.

FINANCIALS

Covering significant funding stories, deals and partnerships, and investment opportunities

• Cerebras Systems solidified its status as the premier challenger to the GPU status quo by closing a massive $1 billion Series H round at a $23 billion valuation. The capital will scale their wafer-scale hardware, aiming to disrupt Nvidia’s dominance by offering a radical architectural shift for training the world’s largest AI models.
  

• Waabi secured $750 million in Series C funding to push its "Physical AI" into the commercial fast lane of autonomous trucking. By ditching traditional modular stacks for an end-to-end AI brain, the company plans to transition from pilot programs to a full-scale robotaxi and freight network.
  

• ElevenLabs has achieved the UK’s biggest tech raise in years this week, by securing a massive $500 million Series D round led by Sequoia Capital. The deal catapults the startup’s valuation to $11 billion – more than tripling its worth in just one year – as it pivots from simple voice cloning toward a full-stack "Audio General Intelligence" platform.
  

• CesiumAstro raised $470 million to bring high-performance AI processing to the final frontier of orbital communication. The deal highlights a surging investor appetite for satellite-to-satellite data links that can handle complex computation in space without relying on terrestrial data centres.
  

• Recursive Intelligence commanded a staggering $300 million Series A at a $4 billion valuation, marking one of the fastest climbs in the history of frontier AI labs. The firm is pivoting away from the "bigger is better" data philosophy to focus on native reasoning capabilities that could challenge the current leaders in LLM performance.
  

• Decagon climbed to a $4.5 billion valuation behind a $250 million Series D as the market shifts its focus from chatbots to agentic AI. This investment reflects a growing enterprise demand for AI agents capable of making independent decisions and executing complex workflows rather than just answering queries.
  

• Positron AI achieved unicorn status with a $230 million Series B specifically targeted at solving the skyrocketing power costs of AI inference. Their new Asimov silicon aims to deliver a massive efficiency leap over traditional hardware, targeting the energy bottleneck that currently threatens the expansion of global data centres.
  

• Bedrock Robotics attracted $270 million to bring autonomy to the grit of the construction site, automating heavy machinery for massive infrastructure projects. Founded by industry veterans, the startup represents the latest bridge between frontier tech and the physical world, addressing critical labour shortages in global land development.

JOIN US ON TELEGRAM  📲

Share this newsletter with a friend and if they give us a follow you’ll gain access to our invite-only Big Machines Telegram group. Here you’ll find like-minded individuals such as builders, investors, founders, and creators.

Access is only granted to those who refer the newsletter to a new subscriber, so if you try sign-up with a work email (this tanks our open rate!), we’ll know, you sneaky bastard.

GO ON, GIVE US A SHARE…  

You currently have X referrals.

UNTIL NEXT WEEK, FRIENDS.

This will never not make me laugh (give us a follow!):