Superhuman buys Fathom instead of building its own meeting notetaker

Share
Superhuman buys Fathom instead of building its own meeting notetaker

Two deals on Monday, one small and one very loud: a productivity suite bought its way into the meeting record, and Elon Musk's AI company dropped the hardest of its claims against Apple.

Superhuman acquired Fathom, the Y Combinator-backed AI meeting notetaker, rather than finish building one itself. CEO Shishir Mehrotra said the company had tested an internal notetaker with select users earlier this year and concluded that doing all parts of meeting recording well is far harder than the finished products make it look; terms were undisclosed. Fathom keeps a generous free plan, which took it past 400,000 monthly active users and more than 1 million recorded meetings, on roughly $30 million raised and a last known valuation of $94 million in 2024. The strategic logic is the meeting as a trigger: notes are the input for drafting the follow-up email, updating the record, booking the next call and starting an agent without anyone prompting it, and Mehrotra teased real-time meeting context as the goal. My read is that this is consolidation among venture-backed notetakers, not a land grab — Granola raised $125 million at a $1.5 billion valuation in March on the same bet that transcription alone isn't a business, and Fathom's investors include Zoom's apps fund plus Reddit's Steve Huffman and Twitch's Emmett Shear, which tells you a $94 million exit was always going to be a good day.


X Corp and SpaceXAI moved to dismiss their federal antitrust lawsuit against Apple in Texas on Monday, resolving the accusation that Apple conspired to unlawfully monopolise markets for smartphones and generative AI chatbots by exclusively integrating ChatGPT into Apple Intelligence. The filing gave no reason and did not say whether a settlement was reached; Musk's companies said they will keep pursuing their claims against OpenAI, the remaining defendant. The part worth noticing is which target they dropped: Apple's default-assistant placement was never really at risk from a lawsuit, and the case's own theory — that distribution lock-in, not model quality, decides who wins consumer AI — is now standing unchallenged as the industry's operating assumption. Expect the next fight over default placement to come from regulators or from a device maker that wants its own model in that slot, not from a rival lab.

What to watch: whether Superhuman keeps Fathom's free tier once it sits inside a paid suite — that's the tell on whether meeting data is being treated as a product or as training ground for agents.

Is the meeting notetaker a product, or just the cheapest way to capture the data that agents need? Tell us in the comments.

Read more

Nvidia-backed Firmus pulls its $30 billion IPO after demand collapses

Nvidia-backed Firmus pulls its $30 billion IPO after demand collapses

The biggest AI-infrastructure listing of the year didn't get delayed — it got pulled entirely, after two price cuts in two days found no takers. Nvidia-backed data-center operator Firmus withdrew its Australian IPO on Friday, abandoning what would have been the country's second-largest listing ever after slashing the offer price twice in 48 hours. The company had been marketing shares at A$11 each to raise about $5 billion at a valuation near $30 billion — roughly three times the $10.5 billion

AI inference is redrawing the storage hierarchy

AI inference is redrawing the storage hierarchy

The money has followed GPUs for years, but at GMIF 2026 in Shenzhen the argument was about everything behind them — and the numbers backing it are getting hard to ignore. China's 140 trillion daily token calls are forcing a redesign of the storage stack. Now in its fifth year, the Global Memory Innovation Forum gathered storage makers, analysts and chip designers around one theme: inference, not training, is now the workload that dictates hardware. China's daily token calls had already passed

Cloudflare launches Clef-omni, cuts Clef-flash price by 58%

Cloudflare launches Clef-omni, cuts Clef-flash price by 58%

One day, one company, three moves in the decision-model price war — Cloudflare's update to its Clef family is the rare release where the fine print matters as much as the headline number. Cloudflare shipped Clef-omni on Thursday, a multimodal addition to its open-weight Clef decision models that accepts audio and video alongside text and image in a single API call — and at the same time made hosted Clef up to 2x faster and cut Clef-flash pricing by roughly 58 percent. Clef-omni takes WAV or MP

Meta turned down Amodei's personal plea for compute

Meta turned down Amodei's personal plea for compute

The chip hunt is back in the headlines — and today's edition runs from a boardroom ask at the very top of the AI industry to the economics of putting robot drivers in truck cabs. Dario Amodei personally approached Meta earlier this year to source more compute for Anthropic, and Meta declined — according to the Wall Street Journal. The report lands inside the Journal's larger piece on how desperately the industry is hunting for computing power, and it is the kind of detail that only surfaces wh