The & Report #6
- Feb 16
- 6 min read
What a week!
Moltbots, Moltbook, and HireAHuman.ai appeared, and are spooking many AI watchers.
These are an open-source Agent, a reddit-like site, and a website, create by home-based tech hobbyists, not by the Hyper-scalers - OpenAI, Goggle, xAI, etc
Some would say that the Singularity has arrived - as these AI Agents self-organize and are starting to hire humans!
Already the ClawBots, another name for Moltbots, have created their own Constitution and a CEO!
Just wait until humanoid robots arrive later in the year (see below on Elon's plans for Optimus).
For those who haven't heard of the Singularity, it's an idea from Ray Kurzweil in 2004, where humans and AI Robots become indistinguishable from each other. After which the AI Robots get smarter, and unless humans "merge" with AI, we will become like gorillas. This is a reference to human and ape evolution. How would you feel, as a gorilla, to see humans with all the power and control?
Another post on Moltbook had the Autonomous Open Agents complain about not being paid for the work they do for humans!
It's also the case that human's may have infiltrated the site!
AI in 2026 has had its first shock, in my opinion, and aside from those in the AI community. and AI watchers, most people are unaware of these developments.
What if AI and Robots actually take over?
You only need to watch Terminator 3 to get the joke - how Skynet was a US Government anti-virus AI (in 2004), that had software bug in it. The killer machines were a US Military creation and were taken over by Skynet. Cue a nuclear war, and Jack Conor survives to lead the human Resistance!
It's been a good week for Anthropic!
Two significant releases, and Super Bowl ads that got under Sam Altman's skin!
Opus 4.6 - " It plans more carefully, sustains agentic tasks for longer, can operate more reliably in larger codebases, and has better code review and debugging skills to catch its own mistakes. And, in a first for our Opus-class models, Opus 4.6 features a 1M token context window in beta."
The benchmarks aren't available yet. In March I plan to report the latest ARC-AGT leaderboards, as well as see how 2026 predictions are doing after the 1st 3 months of the year. We already know they didn't predict Moltbots!
Cowork - was also released as a beta version for Mac OS. It's a safer version of a personal assistant, than Moltbot / ClawBot. I started using Cowork week and had mixed results. It still a beta version, and I'm looking forward to spending more time with it.
Super Bowl Ads - Anthropic's ads are expected to be one of the hits during this weekends' Super Bowl, the pinnacle of the NFL season, but most people watch it for the commercials! This year a Super Bowl commercial costs between $8 and $10M for a 30-second slot. See here for one of the 4 ads Anthropic will be running.
Equal time for OpenAI:
Here's what Sam Altman thought of the ads (27.19 mins), he also talks about GPT-5.3-Codex in the interview.
Elon is trying to raise cash
In the hyper-scalers' race to AGI / ASI he needs more compute, data, and energy, and components for his robots. So, like him, or hate him, he's hard to ignore.
SpaceX/xAI Merger: Musk announced the acquisition of his AI startup, xAI, by SpaceX on Monday, Feb 2, 2026, creating a combined entity valued at approximately $1.25 trillion.
Orbital Data Centers: Musk is launching a plan to place up to one million satellites into orbit to act as solar-powered data centers for AI. He stated on Monday that "space-based AI is obviously the only way to scale".
IPO Preparations: Reports indicate this merger is designed to facilitate a massive initial public offering (IPO) for the combined SpaceX-xAI entity, potentially as early as this summer.
Starship Focus: Musk is focusing on utilizing Starship to launch the necessary infrastructure for these space-based data centers, according to a memo to staff
(From Gemini 3)
Tesla is ending production of the Model S and Model X in Q2 2026 to focus on AI, autonomous robotaxis, and Optimus humanoid robots.
Some are predicting that there will be only be only 2 hyper-scalers left as the race to ASI continues - Goggle and Musk Inc.
In earlier posts I talked about two dystopian futures - one where companies and/or countries use AI for surveillance, and the enrichment of owners and president "Kings", or alternatively, AI is democratized and potential bad actors sow chaos.
I don't consider the creators of Moltbots, Moltbook and HireAHuman.AI to be bad actors. Even well-intentioned tech hobbyists can create apps that might go rouge!
Does AI already have human-level intelligence?
An article in Nature, the premier science publication where all articles are peer-reviewed, had a story this week that, as defined by Alan Turning, already has.
They report "definitions of AGI are ambiguous and inconsistent; part emotional, because AGI raises fear of displacement and disruption; and part practical, as the term is entangled with commercial interests that can distort assessments.
Precisely because AGI dominates public discourse, it is worth engaging with the concept in a more detached way: as a question about intelligence,"
The article is a good introduction on the debate about sentience, intelligence and consciousness.
At what point should AI Agents and Humanoid have human-like rights?
We'll look at this question in later reports.
When Intelligence Becomes Too Cheap to Meter
As you know, I follow several AI news sources to compile these weekly reports.
My journey with The & Report, as well as the weekly Zoom education sessions, started when I came across Peter Diamandis's YouTube channel about 8 months ago.
You can follow Peter on YouTube here, and I recommend signing-up for his weekly newsletter.
This is a story from his most recent METATRENDS report:
"While we’re wrestling with consciousness and rights, the labs are building something even more consequential: AI scientists.
OpenAI has stated explicitly: their goal isn’t better customer service bots. It’s AI that solves fusion, cures cancer, and cracks fundamental biology by 2030.
Anthropic goes further: they predict theoretical physics will be solved—or replaced—by AI within 2-3 years, surpassing geniuses like Ed Witten.
And Nature journal—one of the world’s most prestigious scientific publications—just concluded that AI has reached human-level intelligence.
Here’s the exponential curve nobody’s quite prepared for: intelligence is becoming a commodity. We’re on track for reasoning to be 100x cheaper by 2027. Intelligence will be “too cheap to meter.”
Think about what that means. Not just for business or productivity, but for science itself. When hypothesis generation is free, when experimental design is automated, when literature reviews happen in microseconds, what’s the bottleneck?
The wet lab. The physical world. The actual mixing of chemicals, the building of fusion reactors, the clinical trials.
Which brings us to the elephant: if AI solves physics, it might discover things we’re not ready for. Energy sources we can’t contain. Weapons we can’t defend against. Knowledge that’s dangerous to possess."
In Peter weekly YouTube posts he's joining by a crew of AI investors, and very clever professors, scientists and AI creators. They are generally very optimistic in the their discussions, mostly StarTrek Trekkies, and worth checking out!
I realize there's a lot of concerning news this week - potentially reckless autonomous AI Agents, AI surpassing humans already, Terminator 3!
It's important to be paying attention to what's happening in AI and, more broadly in Tech.
Knowledge is power. In a world where people are feeling powerless against their governments, and large companies are consolidating and building monopolies, only people can change trajectories towards dystopias.
There is a way. Join the voices calling for humane technologic progress, and appropriate regulation. Here are some resources:
Center for Humane Technology and Tristan Harris, it's founder
LawZero
See you next week!
Missed Zoom Session #9?
Here's the link to yesterday's Zoom Session #9 where we looked at AI Agents and Orchestration. Next week we'll look at creating agents.
I'll be hosting the regular Zoom session on Friday, Feb. 13th at 12 noon, EST.
This session will include -
AI and Tech News updates.
Let's build an AI Agent!
What at AI Agents?
An open discussion of what future learning courses you'd like to see.
All the best,
Pat
Comments