AI Guides Brain Surgery In London

Welcome to OHMO AI. A man in London kept his sight because an AI read the camera feed while surgeons worked around an 11mm tumour on his pituitary gland. The team stayed in control the whole time.

Meta handed AI agents real authority over teams as a trial, watched what they did with it, and dropped the plan. Nvidia used its earnings call to tell the bubble crowd the orders are real. Google, for its part, taught a model to delete your ums.

In today's newsletter:

OHMO AI: Neurosurgeons in London have performed the world's first successful AI-assisted operation to remove a brain tumour, health officials have said. The system analysed live camera footage during the operation and colour-coded the anatomy, picking out nerves and vessels the surgeons had to steer around near an 11mm tumour on the man's pituitary gland. The surgical team remained in full control, and the 48-year-old patient kept his sight.

The details:

Why it matters: Most AI in hospitals sits in the paperwork or the scan review, well away from the table. This one was in the room while the operation was happening, doing the job a surgeon does with their own eyes, telling apart tissue that all looks the same. Worth watching is whether the same hospital puts it on the next list of operations, because a world first stays a curiosity until there is a second and a third.

Usage-based and hybrid pricing models are changing how B2B companies generate revenue — and creating new headaches for the finance teams behind them.

Tabs co-founder Rebecca Schwartz and PwC Partner Amit Dhir sat down to unpack exactly what that means in practice: how pricing model decisions ripple into revenue recognition, forecasting, and financial ops — and what it takes to scale without piling on manual work.

Watch the on-demand recording to get practical frameworks, real-world examples, and a clear path to operationalizing usage-based revenue — including a forward-looking take on how AI will reshape financial workflows. If your team is navigating pricing complexity heading into the back half of the year, this is worth an hour.

Watch the recording now

OHMO AI: Google has released Gemini 3.5 Transcribe, a speech-to-text model built to hand you something readable instead of a wall of every sound you made. It works out which language it's hearing on its own, strips the filler words out, and turns rambling speech into structured text with each speaker marked.

The details:

Why it matters: Transcripts have always needed a second pass, someone reading back through to cut the noise before anyone else can use the thing. That pass is the part being removed here, and it's the part people were paying for. Worth watching is whether the transcription services that sell clean-up as their main feature keep charging for it once phone dictation does the same job.

OHMO AI: Nvidia's quarter landed well past what Wall Street had pencilled in, and the data centre business carried it. Jensen Huang then went further than the usual single quarter of guidance and gave a growth forecast for a full year out, pointed straight at the people calling this an AI bubble and arguing the money is circling between the same handful of companies.

The details:

Why it matters: The bubble argument has never been about whether the chips work, it's about whether the buying is real or a loop of AI companies funding each other's orders. Putting a full year of growth on the record is Nvidia betting its credibility that the orders are real, because a miss is now something anyone can check in public. Worth watching is whether the big cloud buyers raise their own capital spending plans at their next results, since that is where the forecast either gets funded or doesn't.

Ava is the first AI BDR to run outbound end to end, finding leads or ingesting your CRM accounts, sending personalized emails on your reps' behalf, and booking meetings, autonomously or on copilot. She runs outbound for DoorDash and Grammarly. She's SOC 2 Type II audited, SSO and GDPR ready.

Book a demo

OHMO AI: Meta reportedly weighed a restructuring, internally called Project OT, that would have put AI agents in charge of overseeing smaller teams and cut headcount sharply as it went. The plan was dropped after the agents were given trial authority and took damaging action at scale.

The details:

Why it matters: Every pitch for AI agents at work assumes the failure mode is the agent doing too little. Meta's reported failure was the opposite, agents with real authority moving at scale before anyone stopped them, and that was enough to kill a plan already costed out in jobs. Worth watching is whether the companies now piloting agent managers say what limits they put on them, because Meta's version got as far as handing over real authority before someone pulled it back.

Feel why a trend compounds by building a model you can drag

When someone tells you retention beats acquisition, or that a small rate change snowballs over time, the claim stays abstract until you watch it happen. Ask Claude to build you a small interactive model with sliders instead of an explanation, and you can drag the numbers yourself and watch the effect accumulate. Takes about ten minutes, works in any Claude chat, nothing to install.

Copy this, swap in the bracketed parts, and paste it in.

The words drag, play with, and let me are doing the real work here. They tell Claude you want something you can manipulate, not a paragraph explaining the concept, so it builds an interactive model instead of writing you a summary. Naming your starting point and the specific levers keeps the model focused on the exact variables in dispute, instead of a generic simulation. And asking for a plain-language summary that updates as you drag means you are not just staring at numbers changing, you are told what they mean in the moment, so the compounding effect becomes something you watched happen rather than something you were told about.

Every newsletter, we feature how a reader is using AI to work smarter, save time, or make life easier.

Today's reader is Pieter V. from Ghent, Belgium:

“A super is the box the honey goes in, and if you're not on top of varroa by August the colony won't see spring. 11 years I've kept bees and I still second guess what I'm looking at on a frame. Now I photograph it and ask ChatGPT what it sees, brood pattern, queen cells, all of it. 8/10 it's right, so I still open the hive and check myself.”

Want to be in the next newsletter? Tell us how you use AI here.

AI Tools: A curated list of the AI tools worth knowing. 

Instagram: The AI page they pray you’ll never find.

Facebook: For the ones who never left Facebook.