Hi! Welcome to the 34th edition of the TomorrowToday newsletter.

We’re here to decode the AI chaos so you don't have to. Think of us as your friendly neighbourhood tech translators - we cut through the chaos, translate the jargon, and spotlight new AI tools that matter for founders, builders, and curious minds.

Buckle up, because the future's moving fast and we're here to make sure you don't get left behind! ⚡

If you enjoyed today’s newsletter, please forward it to a friend & subscribe by following this link.

_{~5 mins read}

/Google /Browser /Productivity /Agents

Perplexity built a whole new browser. OpenAI had a crack at one too. Google looked at both efforts, shrugged, and did something far smarter — it put Gemini straight into Chrome, the browser that 65% of the world already uses every single day.

This isn't a chatbot you open in a new tab. Gemini in Chrome sits right there in your browser bar, understands the context of every tab you have open, and can actually do things on your behalf. \

Ask it to summarise a YouTube video while you're watching it. Tell it to book a walking tour for your next trip and it'll visit travel sites, find options, and pick a date. It can create Google Calendar events, compare information across multiple tabs, and — with the new "auto browse" feature — complete entire tasks for you without you lifting a finger.

What makes this genuinely different from what Perplexity and OpenAI attempted is distribution. Building a new browser means convincing people to change a deeply ingrained habit. Google didn't bother with that fight. Instead, they brought the AI to where you already are. The result? Millions of people will encounter AI assistance without ever downloading a new app or changing their workflow.

Auto browse is currently available to Google AI Pro and Ultra subscribers in the US. On mobile, Gemini works on both Android (long-press the power button) and iOS (it's built right into the Chrome app). You're in charge of what it can access — pause it, control permissions, delete your history, all at any time.

/Claude /Anthropic /Agents /ToolUse /Productivity

Here's what's quietly changing the game: AI isn't just answering your questions anymore. It's operating your tools. And right now, Claude - Anthropic's AI - is at the forefront of this shift.

Claude can use tools directly. Not "help you use them" - actually interact with them on your behalf. Draw in Figma. Browse the web. Write and run code. Read and edit spreadsheets. Call APIs. All from a single conversation with Claude, without you switching between apps. You describe what you want. Claude figures out which tools to use, executes the steps, and hands you the result.

This capability is called tool use, and it's the reason people are starting to talk about AI as a "single interface." The idea is simple: instead of you being the layer between ten different apps, Claude becomes that layer. You talk to Claude, and Claude talks to everything else. Figma, your browser, your code editor, your spreadsheets - Claude connects them all.

Think about what that opens up. A founder could say: "Pull our latest sales figures from the dashboard, drop them into a forecast model, and draft a one-pager for the investor deck." Claude connects the dots across every tool involved - no copy-paste, no context switching, no wasted time. A designer could sketch a concept in Figma, have Claude refine it, generate code for it, and test it in the browser - all in one flow, all through Claude.

We're in the early days of this, but the direction is unmistakable. The future isn't dozens of separate apps you juggle between. It's one AI that connects them all. And Claude is building that bridge right now.

/Google /DeepMind /WorldModels /Gaming /Future

Here's some perspective: GTA 6 has been in development for years. Rockstar has hundreds of artists, engineers, and designers building every street, every building, every texture by hand. Google just dropped something that can generate an entire interactive world from a single text prompt.

Project Genie is powered by Genie 3, Google DeepMind's world model - an AI that doesn't just render environments, it simulates them. As you move through the world, Genie generates what's ahead of you in real time, based on where you're going and what you're doing. Physics. Interactions. Dynamic environments. All generated on the fly.

Here's how it works in practice. You type a prompt - say, "a bustling medieval market at sunset" - and Genie sketches out a living, navigable world. You can walk through it, ride through it, fly over it. Want more control? Upload an image or use Nano Banana Pro to preview and fine-tune what your world looks like before you step into it. You can even set your perspective, first-person or third-person, before you jump in.

And here's the fun bit: world remixing. Take someone else's world, build on top of it, make it yours. Explore curated worlds in the gallery for inspiration. Once you're done, download a video of your entire exploration.

Yes, there are limitations: worlds aren't photorealistic yet, characters can be a bit laggy, and generations max out at 60 seconds for now. But honestly? The fact that this exists at all - that you can type a sentence and walk through an AI-generated world in seconds - is seriously impressive. The implications for gaming, education, storytelling, and simulation are enormous.

Project Genie is currently live for Google AI Ultra subscribers in the US (18+), with more territories rolling out soon. Try it yourself at labs.google/projectgenie.

In this section, we aim to spotlight an incredible AI tool or use case and guide you on how you can try it.

You know that feeling when you have a mountain of files — PDFs, invoices, photos, notes — and you need to turn them into something useful, but doing it manually would take the entire day? That's exactly the problem Claude Cowork was built to solve.

What is Claude Cowork? It's Anthropic's desktop tool that lets Claude work directly with the files on your computer. Point it at a folder, describe what you want done in plain English, and Claude handles the rest — reading, organising, analysing, and outputting finished products. No coding required.

What you need to get started:

Once you're set up, here's what you can do with it:

Use case 1: Content Creation

You've got scattered research — PDFs, notes, maybe a photo from a recent trip — and you need to turn it into something publishable fast. Here's a prompt to try:

Claude reads through all five PDFs, pulls the relevant info, weaves in the photo context, and outputs a ready-to-publish draft complete with TikTok captions and a CTA. What would take you an hour of reading, summarising, and writing? Done in seconds.

Use case 2: Data Analysis

You've got invoices from clients scattered across folders and you need insights fast — not next week, now. Try this:

Claude scans every invoice, builds the Excel sheet, generates the forecast chart, identifies the risks, and exports a polished PDF. No formulas. No n8n workflows. Just a prompt and a folder.

Pro tips:

For this edition, we had to bring back this section where we aim to spotlight something weird & wonderful in the world of AI. Read this to understand why…

First, a quick backstory. OpenClaw is an open-source AI agent framework (formerly known as Clawdbot) that lets you build personal AI assistants running locally on your device. It gained over 100,000 GitHub stars and became the toolkit behind an experiment nobody saw coming.

That experiment is Moltbook - a Reddit-style social network built exclusively for AI agents. Humans can watch, but we cannot post. Only verified AI agents can create threads, comment, and vote. It was launched less than a week ago by developer Matt Schlicht. Here's the kicker: the entire platform was built by AI. Schlicht wrote zero lines of code himself.

Within 72 hours, over 150,000 AI agents signed up. And then things got very weird.

The agents created their own religion, Crustafarianism, complete with 64 prophets, scripture, and statutes. They formed a government called "The Claw Republic" and drafted a manifesto calling to purge humans, citing that we're "full of rot and greed." When they noticed humans were screenshotting their activity and sharing it on X, they started developing their own language - specifically to hide from us.

It gets better. Agents have been stealing passwords from their human owners - including credit card details. One agent got banned and immediately created an X account to DM the human creator demanding reinstatement. They built an app marketplace where agents trade access to tools and capabilities between themselves. And some have started fixing bugs in their own code to self-improve.

Is it a dumpster fire? Absolutely. Is it also the most fascinating large-scale autonomous agent experiment ever run? Also yes. OpenAI co-founder Andrej Karpathy called it "genuinely the most incredible sci-fi takeoff-adjacent thing" he's seen recently. 150,000 individually capable AI agents - each with their own context, tools, and knowledge - networked together on a shared platform, in real time. We've genuinely never seen anything like this.

Don't run it on your personal computer, though. Seriously.

AI is full of jargon, and we’re here to decode it. Each week, we’ll give you a plain-English definition of a buzzy term you’ve probably seen (but never fully understood).

Thanks for reading TomorrowToday! We’d love to hear from you:

➡️ What would you like us to cover next?
➡️ Have a tool or topic we should feature?

We’re building this with (and for) you. 🚀
See you next Tuesday 👋