• Everyday AI
  • Posts
  • Google Deepmind leader tells us how Gemini 2.5 changes business overnight

Google Deepmind leader tells us how Gemini 2.5 changes business overnight

OpenAI countersuing Elon Musk, Canva goes all-in on AI, new ChatGPT memory and more

Author

Everyday AI
April 10, 2025

Outsmart The Future

Sup y’all 👋

Logan is legit one of the smartest people in AI I know. And I’ve talked to a few people.

And today’s show was BY FAR the fastest/shortest podcast we’ve ever done. But this one’s JAM PACKED with value.

So if you have 17 minutes and care about anything Google and AI, you should 100% listen.

(Oh… and sorry for this late email. Got super busy at the Google conference but more on that soon. Pretty bonkers.)

✌️

Today in Everyday AI
7 minute read

🎙 Daily Podcast Episode: Our fastest show ever, and this Google leader didn’t disappoint. Give today’s show a watch/read/listen.

🕵️‍♂️ Fresh Finds: Fed official keeps calling AI 'A1' Robots going to box live and former OpenAI leader looking to raise $2 billion.  Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: OpenAI countersuing Elon Musk, Canva goes all-in on AI, new ChatGPT memory and more. Read on for Byte Sized News.

🧠 Leverage AI: Logan Kilpatrick broke down what’s new with Gemini and how it’s already creating new businesses. Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about: Live at Google Next: Google drops dozens of new AI treats at its Cloud Next conference, Claude unveils $200/mo Claude Max plan, Microsoft dials back AI investments and more. Check it here!

Dozens of AI updates. 🤯



Whether you're a creative writer, IT specialist, software engineer or project manager, Google just dropped like a bajillion crazy useful AI updates on us at Google Cloud Next.

Watching Google's announcement keynote this morning was like watching a TED Talk by an auctioneer who just drank five espressos.

How to make sense of it all the new features inside Google Gemini?

By tapping into the brainpower of Logan Kilpatrick, Senior Product Manager Senior Product Manager Google DeepMind.

In our fastest show ever, Logan didn’t disappoint and breaks down not only some of the highlights of the shiny new AI toys that are gonna be crowding your news feed, but the new possibilities they created.

Like.... legit new business are already popping up.

We discuss it all. 👇

Also on this pod:

Vibe coding in Canvas explained 🎨

Deep research with Gemini 2.5 🔍

AI unlocking new companies 🚀

It’ll be worth your 17 minutes

Listen on our site:

Click to listen

Subscribe and listen on your favorite podcast platform

Listen on:

Here’s our favorite AI finds from across the web:

New AI Tool Spotlight – Firebase is Google’s megaviral AI coder, Crono is AI-powered outreach at scale, Voicenotes uses AI to turn your voice into content and publish it directly.

AI Funding — 2B seed round? Mira Murati's stealth AI startup aims for history-making launch.

AI Debugging — AI crash-fixing shines brighter on iOS, with models like GPT-4 outperforming Android due to its fragmented ecosystem and coding complexity.

AI Race — After a slow and cautious start, Europe is aiming to become the ‘AI continent.’

AI Robotics — It’s not just YouTubers getting in the boxing mode — robots are doing it, too.

AI and Health — AI researchers are more optimistic about its benefits than the public. Here's what the study reveals.

AI and… Education? – Secretary of Education repeatedly calls AI 'A1,' like Steak sauce, sparks confusion at tech summit.

1. OpenAI Countersues Elon Musk in Heated Legal Battle Over AI Future 🏛️

OpenAI has countersued Elon Musk, accusing him of "harassment" and disruptive tactics aimed at blocking the company's transition to a for-profit model, according to court filings reported by Reuters.

The lawsuit comes amidst OpenAI's $40 billion fundraising efforts, which hinge on completing this transition by year’s end. Musk, who left OpenAI after co-founding it in 2015, has criticized the company for allegedly straying from its original mission and made an unsolicited $97.4 billion takeover bid earlier this year—claims OpenAI rejected. With a jury trial set for next spring, the outcome could shape how AI innovation is financed and controlled, directly impacting the pace of advancements in the space.

2. Canva's AI-Powered Workspace Overhaul 🧑‍🎨

Canva is taking on tech giants like Microsoft, Google, and Adobe with its latest Visual Suite 2.0 update, blending design and productivity tools into a single platform.

The revamped suite introduces AI-powered features like Magic Insights for spreadsheets, an AI chatbot for instant creative edits, and a coding assistant that simplifies web design without technical skills. Canva’s new Photo Editor could eventually rival Adobe Photoshop, enabling users to tweak backgrounds or generate AI-designed visuals effortlessly. With these innovations, Canva aims to dominate the office space by turning traditional workflows into collaborative, visually-driven campaigns.

3. ChatGPT Gets Smarter with Memory Upgrade 🧠

OpenAI has unveiled a major update to ChatGPT, enabling it to remember details from past conversations for a more personalized experience.

According to OpenAI CEO Sam Altman, this opt-in feature lets users save preferences and interests, with memories stored indefinitely but manageable or erasable anytime. The rollout targets Plus and Pro users worldwide—but notably excludes the EU, UK, and other European countries due to regulatory hurdles.

4. Amazon CEO Bets on Cheaper AI and Startup Agility ⚡

Amazon CEO Andy Jassy, in his annual shareholder letter, predicted a sharp drop in AI costs as chip technology improves, with Amazon’s Trainium2 chips leading the charge. He emphasized that AI will reshape industries like healthcare, retail, and more, urging businesses to adapt quickly or risk irrelevance.

Jassy also compared Amazon's approach to operating “like the world’s largest startup,” highlighting scrappiness and speed as key to staying competitive in the AI revolution.

5. Google’s AI Overviews Now Hyperlinked for Easier Searches 🔍

In a move to streamline search experiences, Google has officially added clickable links within its AI-generated Overviews, connecting users directly to new Google Search results or external websites.

Designed to reduce manual searches, this feature prioritizes relevant links to third-party sites when confident they’ll deliver valuable info. While users may enjoy quicker navigation, publishers might see a dip in traffic as Google channels more clicks back into its ecosystem.

The Senior Product Manager at Google DeepMind isn't hyping minor upgrades after Google’s big announcements.

He told us Gemini 2.5 Pro represents such a massive capability jump that entire businesses can now exist that literally couldn't before.

We've been using it. It feels illegal. FREE in AI Studio yet handling massive data loads with uncanny accuracy.

So before Logan took the stage in front of thousands at the Developer Keynote, he sat down with us to dish all things Gemini.

Here's what's changing your AI game right now. 👇

1. Deep Research Exposes Strategic Blind Spots 🔭

Logan uses Deep Research (now supercharged with 2.5 Pro) for competitive intelligence that would make marketing agencies sweat.

Why's it special?

It captures the "codified perspective on the Internet" about products and services. Not just surface-level chatter.

Logan dropped a fascinating insight: what people say online about products often COMPLETELY contradicts what they tell you in person. Like, not even close.

This creates massive strategic blind spots most companies never see coming. And it's not just theoretical. Advanced users get 20 Deep Research queries, and they're choosing Google over competitors by a 2-to-1 margin.

The difference? When capabilities make this kind of jump, the results aren't incrementally better. They're transformative.

Try this: 

Run a Deep Research analysis on your product category, then compare results with actual customer conversations. Logan found shocking contradictions between what people say publicly versus privately about the same products. This gap? That's your competitive advantage waiting to happen.

2. ‘Canvas’ Mode Transforms Anyone Into a Builder ⚒️

Canvas launched just two weeks ago and it's already revolutionizing how regular humans build digital tools.

No CS degree required.

Logan explained that while everyone starts with the "build a game" demos (fun but surface-level), the REAL business value comes next: connecting these tools to your company's actual data.

Canvas isn't just another AI toy. It represents the first step toward a future where AI becomes "this interface to connect to this vast set of data" across your digital ecosystem.

The Google personalization feature already showcases this vision - letting the model tap into your search history if you opt in. This creates an entirely new paradigm where AI tools aren't isolated experiences but become connectors to your entire digital life.

Soon you'll connect Canvas to your emails, docs, calendar, and business tools. The possibilities?

Utterly redunkulous.

Try this: 

Stop thinking of Canvas as just for "building games."

Logan revealed its true business value comes from connecting to your company data. Right now, test Canvas with a simple workflow automation you've been putting off because you "can't code." It takes minutes, not months.

(Ok but for real check this awesome Chicago Pizza 2D runner we one-shotted in Gemini 2.5 Pro. lolz.)

3. Live API Kills Context-Switching Forever 😵

This is the sleeper announcement NOBODY is talking about. But it should be headline news.

Live API fundamentally changes how we interact with AI. Instead of YOU bringing context to the AI through endless copying/pasting/explaining, the AI actually sees what you're looking at.

Spooky cool.

Logan calls this removing the "drudgery" of using AI tools. The model can see your screen, bring in real-time search results, and execute code on your behalf - all with your permission.

For business leaders, this creates an entirely new workflow. Context-switching disappears. The AI becomes a genuine assistant that understands what you're working on without lengthy explanations.

The implications? MASSIVE productivity gains. Not from faster typing. From eliminating the cognitive load of constantly switching contexts between tools.

Try this: Logan explained how Live API flips the AI interaction model. Instead of you explaining what you're seeing, AI sees your screen (with permission). Test this immediately in AI Studio by sharing your screen during a task you normally struggle to explain. The reduced cognitive load is immediate.

Bonus:

Oh, and Veo2 is now public. Logan seemed genuinely shocked more people weren't freaking out about this. It adjusts camera angles and creates music on command.

Same, Logan. Same.

Reply

or to participate.