AI Academy
Posts
🤖 Apple Predicts Google's Search Sunset

🤖 Apple Predicts Google's Search Sunset

Plus: Gemini's Video Wizardry and Deep Research New Features

May 15, 2025

Hello AI Enthusiast,

Apple's comments about the future of search just sent Google's stock on a rollercoaster ride. Meanwhile, Google's showing what's possible when AI meets video content, and OpenAI keeps making Deep Research more practical with code integration. Let's look at what's really happening behind the headlines.

The Big Picture 🔊

Apple Exec: AI Will Replace Traditional Search Engines

Apple's services chief Eddy Cue testified that AI search engines will eventually replace traditional search engines like Google. He plans to add AI services from OpenAI, Perplexity, and Anthropic as Safari search options. This testimony came during the remedies phase of Google's antitrust case, causing Alphabet shares to drop over 7%. Alphabet responded with a statement claiming continued search query growth, including from Apple devices.

Gianluca MauroFounder and AI Rockstar

This is a classic platform shift moment. Many of us already use tools like Perplexity instead of Google for certain queries - why hunt through links when an AI can just give you the answer? Google's nervous-sounding response tells us everything: "Don't worry, our search traffic is still growing!"

Meanwhile, Apple collects $20 billion annually from Google while quietly preparing for a different future. The question isn't if AI will change search, but who'll win this high-stakes game of musical chairs.

And here's something we're excited to share - people are discovering our programs through AI recommendations! One participant joined us after ChatGPT suggested our bootcamp when he asked about AI training. As search patterns evolve, our quality content is naturally rising to the top of AI suggestions. Not getting AI recommendations yet? Hear what others are saying about our bootcamp and join us before the next cohort!

Google's Gemini 2.5 Models Tackle Video Analysis

Google has launched Gemini 2.5 Pro and Flash with advanced video understanding capabilities. The models can transform videos into interactive applications, generate animations, and pinpoint specific moments in content. Google claims it can process up to 6 hours of video, available through their AI Studio and API offerings.

Gianluca BelloniCMO and Marketing Nomad

We've been testing this feature ourselves and it's genuinely impressive. Feed it a complex physics video and it creates interactive simulations that make abstract concepts tangible.

This approach of experimental innovation (similar to how NotebookLM started as a side project) is exactly what we teach in our AI Agent Bootcamp: the best innovations often come from building quick prototypes to solve real problems.

OpenAI Upgrades Deep Research

OpenAI has added two practical features to Deep Research: GitHub repository integration that lets the AI analyze code and PRs to create cited reports, and the ability to export research as well-formatted PDFs with tables, images and linked citations. The export works for both new and past reports via a simple "Download as PDF" option.

James VarnhamCEO and Rainmaker

We've tested Deep Research on real problems and it's impressive - doing in 90 minutes what would cost thousands in professional fees. It's not perfect (we've caught it mixing information), but the implications are clear: many service industries are facing disruption.

This doesn't eliminate professionals completely, but changes when you need them - use AI for the exploratory research phase, then bring in experts for their specialized knowledge and judgment.

Bits and Bobs 🗞️

Mistral AI's new model, Mistral Medium 3, offers competitive performance, especially in coding and STEM tasks, at a lower cost than leading models.
Notion and Replit have teamed up to let users build apps using Replit for the frontend and host content on Notion as the backend.
A recent study showcases a new automated system using Gemini models that iteratively refines prompts to simplify complex texts without losing key details.
OpenAI has announced that Fidji Simo will join as CEO of Applications, enhancing their expansion in delivering AI products globally.
OpenAI has launched "OpenAI for Countries," an initiative within the Stargate project, to assist global nations in building democratic AI infrastructure.
A new benchmark called HealthBench has been introduced to better evaluate AI systems in healthcare, ensuring they are useful and safe.
Aurora has launched commercial driverless truck services in Texas, in partnership with Uber Freight and Hirschbach Motor Lines.

From Our Founder’s Channels 🤳

Check out Gianluca's latest TikTok where he breaks down Zuckerberg's AI vision and what it really means when Meta wants to leverage all that personal data they've collected over the years.

@gianluca.mauro
Zuck’s vision for AI is so dangerous #Ai #learnontiktok #artificialintelligence #business #machinelearning #product #ux #entrepreneur

LOLgorithms 😂

They have a point.

That's a wrap on our newsletter! Before you go, here’s a quick recap of our offerings:

AI Academy Membership: Get 12 months of access to all our cohort-based programs, live webinars, on-demand courses, and tutorials.
AI Agent Bootcamp: Accelerate processes and solve business problems by mastering prompts and building AI Agents, without coding.
Practical Introduction to ChatGPT: A free course on using ChatGPT confidently, understanding its workings, and exploring its potential.
Customized Corporate Training: Equip your team with the skills they need to unlock the potential of AI in your business.

Catch you next week! 👋