The skinny on what's happening in AI - straight from the previous live session:

Highlights

🌐 OpenAI Launches Atlas Browser - Google's Worst Nightmare

Discussed at [00:00:26]

OpenAI surprise-launched Atlas, a Chrome-based browser that replaces the URL bar with ChatGPT. Available only on Mac with M-series chips. The magic: it remembers every website you visit, can act as your agent, and pulls Google search results WITHOUT showing Google's ads.

Kyle's take: This is a full frontal shot at Google. The URL bar is now a chat bar - you type something, it queries ChatGPT by default. You can also search like in a normal browser but that’s a secondary step. Chat is primary. A new paradigm for using the internet.

Most aggressively when I did search Atlas scraped Google results and showed them directly in Atlas WITHOUT ads. Google gets zero revenue while still paying for indexing. ChatGPT is just … using the results without doing the work? Sneaky!

And what happens when OpenAI adds their own ads? Google’s hard work but OpenAI’s revenue? This could be a legal fight.

The agent mode is painfully slow - like watching your nan use a computer ("click there... no, there... come on"). Found me £400 flights to Taiwan but watching the cursor crawl around was torture. BUT…get a handful of agents working in parallel? Or set them on tasks overnight? Then we’re talking.

Personally, I refused to give it my passwords - they haven't proven they can secure this data.

Source: OpenAI Atlas launch / chatgpt.com/atlas

💰 Google's Impossible Pivot: 57% of Revenue Under Threat

Discussed at [00:21:10]

Why exactly is Atlas such a big problem to Google? It’s just search that’s threatened right? Yup…BUT, and it’s a big but: Search is Google’s cash cow.

Google makes 57% of its $350 billion revenue from search ads. Cloud services ~12%. YouTube? Only 10%. Chrome controls 65-70% of browser market share and Google searches amount for 90% of global searches.

Now OpenAI wants both.

Kyle's take: Google's in an impossible position. They need to replace 60% of $350 billion revenue - that's $200+ billion at risk.

They can't pivot to AI fast without killing their cash cow. But if they don't pivot, OpenAI eats their lunch.

Every AI feature Google adds (AI overviews, AI search mode) cannibalises their own ad revenue. Which is better than giving it to OpenAI sure but will it generate enough cash? Right now that’s a total unknown. None of the AI companies are fully monetising and none from ads in their results.

It's like turning a cruise ship while it's sinking. If anyone can pull it off, it's Google - they're smart, nimble for their size and have some of the world’s best engineers.

But this is existential. Ironically the DOJ just barred them from search exclusivity deals, and now OpenAI provides the "alternative" the government wanted. I don't envy anyone at Google today. It’s going to be a complex few years.

🛠️ Google Quietly Releases AI Studio Build - Terrible Timing?

Discussed at [00:23:00]

Google released their own vibe coding tool yesterday - AI Studio Build. Like Lovable/Bolt but free. And nobody noticed because Atlas dominated the news. Bad timing….

Kyle's take: Classic Google - amazing product, zero marketing. I don't even know what it's properly called. Built a YouTube thumbnail generator in 5 minutes - pulls my newsletters, generates images with nano-banana, overlays text that actually works (rare for AI, believe me!), and adds my reaction photos with sentiment analysis.

The AI images look like slop right now but that's fixable with better prompts. It's also FREE unlike Lovable/Bolt/Replit.

Definitely go and have a go!

Source: Google AI Studio Build 

Want the full unfiltered discussion? Join me tomorrow for the daily AI news live stream where we dig into the stories and you can ask questions directly.

Streaming on YouTube (with screen share) and TikTok (follow and turn on notifications for Live Notification).

Audio Podcast on iTunes and Spotify.