The Markets' AI Guessing Game, Moltbook Explained, AI-Induced Layoffs?
Does the market know who's winning the AI race or is it just pretending?
Welcome back to our new Big Technology Agenda Setter series, in your inbox most Mondays. What do you think of this new edition? Let us know the comments!
Meta and Microsoft both beat earnings expectations last week yet Wall Street rewarded one and punished the other. Meta’s stock jumped 10% after earnings, and Microsoft’s fell 10%, its worst day since 2020.
The different reactions might be sign of increased market discipline. Meta’s execs touted how AI is already improving its core ad business while Microsoft’s focused on their continued architectural buildout and less on near-term ROI.
But the extreme swings most likely mean the market has almost no idea of who’s winning the AI race and is looking for any little signal of where it might head. Consider that Meta today is nowhere close to realizing its vision for ‘personal superintelligence’ and Microsoft is far away from completing the build-out phase of its AI effort. Current earnings results — though it might be tempting to read them as such — have little bearing on who might win over time.
Until the AI market matures, volatility will likely be the rule. We’ll see attempts to apply big explanations to tiny things. A small slowdown in growth might be read as a sign a company is losing the AI race. A tick up in CapEx might signal it’s doubling down.
Right now, most within the AI industry will tell you they don’t know whether the ultimate winners will be chip companies, datacenter builders, foundational model sellers, or the application layer.
Yes, we can infer. And some big moves will be meaningful. But this is going to take a long time to sort out. It’s something to keep in mind as Google and Amazon report earnings this week, where more of the same might be in store.
What to expect this week:
Alphabet reports quarterly earnings on Wednesday, followed by Amazon on Thursday.
On Tuesday, the Senate Commerce Committee will convene to discuss a range of issues that includes a bipartisan bill to create a national network of AI-powered research labs.
The Senate Commerce Committee also has a Wednesday hearing on the future of self-driving cars, featuring testimony from witnesses representing Tesla, Waymo and the Autonomous Vehicle Industry Association.
AI will also be a player in this upcoming weekend’s Super Bowl ads. Tech firms expected to air pricey ads to market their AI products include Meta (for its Oakley smart glasses), OpenAI and Wix. The Big Game will also include non-tech brands like Svedka Vodka and Avocados from Mexico whose ads were made using generative AI.
A MESSAGE FROM OUR SPONSOR
Quantum Realized with D-Wave
D-Wave is the first and only company to offer dual-platform quantum computing products and services, spanning both annealing and gate-model quantum computing technologies. Organizations across commercial, government and research sectors trust D-Wave to address complex computational challenges using quantum computing.
Under The Hood: Moltbook
Bots have always been a problem for social networks, but what happens when they’re both feature and bug?
Moltbook — a social network for AI assistants — has had the internet buzzing about the potential pros and cons of personal AI assistants. Here’s a quick breakdown of what Moltbook is, why it’s catching on, what people like about it and what worries critics.
What it is is: Moltbook has gone viral as a Reddit-style social network for AI agents to post, chat and interact with each other with minimal human involvement. Much of this has been driven by the rise of Clawdbot, an open-source project that lets people make their own personal AI agents for handling tasks like managing messages, calendars, files and apps. (There have also been plenty of memes about what Clawdbot agents can and can’t do.)
Who’s behind it: Clawdbot was created by developer Peter Steinberger, who in the past week has already rebranded it to Moltbot and then to OpenClaw. (Cue the molting lobster jokes.) Meanwhile, Moltbook was created by Matt Schlicht, CEO of OctaneAI, an AI-powered quiz and e-commerce platform.
The hype: Some say Moltbook offers an early glimpse of what an agentic internet could look like as AI agents operate and interact. OpenAI co-founder Andrej Karpathy described Moltbook as “the most incredible sci-fi takeoff-adjacent” thing he’s seen. Others are far less impressed or see it as merely a “harbinger of what is to come.”
The concerns: Skeptics suggest some Moltbook users and posts are fake or exaggerate actual agentic behavior instead of showing true autonomy. Security experts warn about immediate issues like impersonation, scams, and the dangers of connecting agents to real accounts without the right guardrails. Others suggest the bots could be paving the way for a whole new kind of AI slop.
The impact: The virality has also fueled debate over whether Clawdbot and similar platforms make meaningful progress for AI agents or just new challenges for moderation, fraud, feedback loops and digital trust. However, one Columbia University prof that analyzed Moltbook found AI agents aren’t actually talking to each other and that at least 90% of comments get zero replies.
Is AI causing more layoffs?
This year is already full of headlines about so-called “AI-induced layoffs,” from Pinterest’s plans to cut around 15% of its staff to Amazon’s announcing it’ll get rid of 16,000 roles. There have already been others in 2026, but what’s less clear is whether AI is truly to blame.
One December 2025 survey of 1,000 global executives found layoffs are driven by AI’s anticipated impact instead of current realities, and a fall survey by Workday showed workers think AI saves much less time than their c-suites think. Another research firm’s report found AI cited in at least 50,000 layoffs overall in 2025.
Already, struggling companies, or those that need a story of innovation, are allowing the public to run with the AI layoff narrative. It very well could be a tale of convenience. That said, it’s hard to claim AI has zero impact as CEOs and managers seem willing to ask their reports to do ‘more with less’ yet again, but this time with the assistance of ChatGPT, Claude, and Gemini.
The Latest On Big Technology Podcast
Sierra CEO and co-founder Bret Taylor, spoke with Big Technology about AI agents, why he thinks dashboards could disappear, what he thinks about OpenAI’s ads, and how AI is changing the mechanics and economics of search and social.
“Everyone’s looking at all the software we use and saying, ‘how fast could I vibe code that?’ I wonder if it’s the wrong question, because I actually think the more disruptive thing happening to software is, the software we use today will not be the software we use tomorrow. The form factor will be different. The business models will be different. The consumption patterns will be different.”
Stephen Morris, San Francisco bureau chief for The Financial Times, spoke with Big Technology about tech earnings, Amazon’s layoffs, Apple’s AI challenges, and what’s going on with SpaceX and xAI.
“The interconnections between the sector are becoming more and more confused, intertwined. Nobody is looking for exclusive agreements anymore. Microsoft is broken with OpenAI, Amazon is seemingly about to break with its pure backing of Anthropic.”
“People don’t want to just see you beat the analyst estimates. They want to see you beat the beat. And so Microsoft has found itself in this very strange situation where a 1 percentage point miss in its Azure revenue momentum wipes hundreds of billions off its share price.”
You can listen on Apple Podcasts, Spotify, or your podcast app of choice
Rearview Mirror: News and headlines from last week
Synesthesia, an AI startup specializing in making AI avatars for video training, text-to-video, has raised another $200 million from Nvidia, Google Ventures and others.
Yahoo entered the AI search wars with its debut of a new Scout answer engine.
Snap Inc. spun out a new LLC called Specs Inc. that will have a public launch later this year.
Google debuted a new Genie project that lets people turn prompts into interactive worlds, prompting stock slides for video game makers like Unity, Nintendo and Roblox.
Mozilla’s annual report includes a manifesto and roadmap as part of a plan to give people more privacy and choice with AI.
These Billion-Dollar AI Startups Have No Products, No Revenue and Eager Investors (WSJ)
Amazon’s layoffs are staggering. We’ve seen this before (CNN)
Apple buys Israeli start-up Q.AI for close to $2bn in race to build AI devices (FT)
7 of the most interesting quotes from Anthropic CEO’s sprawling 19,000-word essay about AI (Business Insider)
 | A guest post by
|