- Innovation Network Newsletter
- Posts
- McKinsey Quietly replaced 5000 consultants with AI | Ep. 32 🏄♀️
McKinsey Quietly replaced 5000 consultants with AI | Ep. 32 🏄♀️
The newsletter to thrive in an exponential world
Google dropped 100 announcements at I/O 2025, but three signals matter most: AI that sees, hears, and acts autonomously.
Meanwhile, ex-Apple designers chase $6.5 billion dreams with mysterious necklaces, while one guy builds million-dollar businesses using five AI tools from his bedroom. From hospital simulations running at 100x speed to entire countries getting ChatGPT access, the acceleration continues.
This week, we explore why Google's video AI breakthrough changes everything about content creation, examine the psychology behind voice manipulation, and reveal how AI is already reshaping million-dollar consulting firms.
We'll also show you why the job market transformation is happening faster than anyone predicted, and preview the strategic masterclass that separates signals from noise.
And last but not least, Patrick’s new book Golven van Overvloed (Waves of Abundance in English) is now available for pre-order! Order it now to dive into five waves of change that will reshape everything.

Let’s keep riding this tech wave together!
And don’t forget to follow us on socials and share share share all the innovation goodies!
Cheers,
Patrick, Nikola & Aragorn 🚀
Google's Revolution
Google's Veo 3 just became the first AI video generator that creates synchronized audio alongside visuals. Eight-second clips now feature people having conversations, comedians telling jokes, and influencers creating content. All completely AI-generated.
"Despite the fact that I know this was going to happen, it still blows my mind,"
Google's video generation breakthrough is rolling out to additional countries, but European users remain locked out of the AI content revolution. The workaround remains unchanged: Google AI Pro subscription plus US VPN connection unlocks access to the platform that's already transforming how creators produce professional content.
"After lunch, the Americans come online and you have to give the prompt 10 times before it actually starts generating anything because America is just completely overwhelming Google's capacity," Aragorn notes from his testing experience. Morning sessions in European time zones offer the best performance before US demand crashes the servers.
Why does it matter?
We're entering the age of infinite personalized content where anyone can become a Hollywood-level producer. The advertising industry will be transformed first, followed by education, training, and entertainment. Organizations that understand this shift from expensive production to intelligent generation will capture massive competitive advantages while traditional content companies struggle with obsolete cost structures.
Project Astra: Google's Jarvis Moment
Google unveiled Project Astra! An AI assistant that doesn't just answer questions but autonomously controls your phone, makes calls, schedules meetings, and navigates apps. The demonstrations show AI downloading manuals, scrolling to relevant pages, opening YouTube videos, and conducting phone calls to order parts.
The technical reality includes important caveats: Google sped up the demonstration footage (marked "2x speed" in small print), indicating current performance is slower than marketed. However, the Model Context Protocol (MCP) integration allows AI to communicate with non-AI native apps through unified APIs, enabling control of existing software without rebuilding everything.
Why does it matter?
This represents the transition from AI as a conversation partner to AI as an autonomous agent. Google is solving the integration challenge that Apple's closed ecosystem makes nearly impossible. Organizations that prepare for agentic AI will gain massive productivity multipliers, while those treating AI as just smarter search will be left behind.
The Voice Manipulation Breakthrough
Hume AI's EVI (Empathic Voice Interface) now allows users to design custom AI voices with specific psychological characteristics. Request a "low, dark voice that sounds mature and professional," and the system generates precisely calibrated audio designed to influence listener behavior. The technology builds on negotiation psychology, lowering your voice and using downward intonation creates authority and compliance.
The level of manipulation that an AI can potentially do even if this is under control is just off the charts
Think about it. The same voice modulation techniques that professional negotiators use to close deals can now be programmed into AI systems that interact with millions of users daily.
Why does it matter?
We're approaching an era where AI doesn't just process information but psychologically influences how humans respond to it. Organizations that understand voice psychology and can ethically deploy these capabilities will shape human behavior at scale. The strategic question becomes: how do you prepare for AI that knows exactly how to sound to get what it wants?
The $6.5 Billion Necklace Mystery

via OpenAI
Ex-Apple designer Johnny Ive sold his company to OpenAI for $6.5 billion to build "the next iPhone" for AI interaction. Early leaks suggest a necklace form factor with camera and audio recording capabilities.
The fundamental challenge remains privacy and social acceptance. Google Glass failed partly because people feared secret recording. A wearable AI that captures continuous audio faces the same resistance—people will modify their behavior around someone wearing such a device, limiting its practical utility.
If they're releasing a necklace.... if they want to fail, that's the way to do it
Why does it matter?
The race to replace smartphones with AI-native hardware is accelerating, but form factor matters enormously. Smart glasses solve multiple problems simultaneously—better audio positioning, natural camera angles, and social acceptance. Organizations betting on the next computing platform need to understand that technical capability alone doesn't guarantee market success.
Claude 4: The Intelligence That Feels Human
Anthropic's Claude 4 doesn't dominate benchmarks but delivers something else that is valuable: natural conversation that feels genuinely intelligent. While technically slower on graduate-level reasoning and math compared to competitors, Claude 4 excels at the human element of AI interaction.
The breakthrough lies in conversational sophistication. Claude 4 asks the right questions at the right moments, maintains context naturally, and provides responses that feel like talking to an intelligent colleague rather than querying a database. This represents a shift from raw computational power to emotional and social intelligence.
Why does it matter?
As AI capabilities plateau on pure performance metrics, the competitive advantage moves to user experience and natural interaction. Organizations that prioritize AI that humans actually want to work with will see higher adoption and better outcomes than those chasing benchmark scores. The future belongs to AI that augments human intelligence rather than replacing it.
Perplexity's Creative Revolution
Perplexity Labs deep research capabilities that rival ChatGPT while adding something unique: complete app generation from single prompts. Request a "futuristic social media platform for 2030," and it delivers a fully functional website with assets, sources, and professional design. The integration combines research depth with creative execution.
I'm very impressed because I already used it this week to do some research and to draft me a chart. The stuff that came out is absolutely on par with ChatGPT's own deep research and even better in some ways
The platform now competes directly with specialized tools like Gamma and Lovable while maintaining its research advantages.
Why does it matter?
The boundaries between research, design, and development are dissolving. Single platforms that can research topics, generate assets, and build functional products will replace entire workflows that currently require multiple specialized tools. Organizations that adopt integrated AI platforms will operate faster and more efficiently than those managing fragmented toolchains.
MCP: The Protocol That Connects Everything
Google and Anthropic are pushing the Model Context Protocol (MCP)—a unified API that allows AI to control any application, even those never designed for AI integration. Instead of rebuilding every app for AI compatibility, MCP creates a translation layer where YouTube, GitHub, databases, and phone apps all speak the same language to AI assistants.
You can make your YouTube talk to the MCP protocol and MCP protocol talks to the AI and that way you can control YouTube with your AI even though YouTube was not built with AI integration in mind
The breakthrough solves the massive integration challenge facing every organization trying to deploy AI across existing software ecosystems.
This represents Apple's nightmare scenario. Their closed ecosystem strategy, built over two decades to control every user interaction, becomes a liability when AI needs universal access.
Why does it matter?
MCP transforms AI from a conversation tool into a universal automation platform. Organizations that adopt MCP-compatible systems will gain autonomous AI capabilities across their entire software stack, while those locked into closed ecosystems will watch competitors operate at AI speed. The protocol war between open integration and walled gardens will determine which companies can actually deploy AI at scale versus just talking about it.
Elon's $300 Million Telegram Masterstroke

XAI paid Telegram $300 million to integrate Grok into the chat platform, gaining access to 1 billion users plus their conversation data for training.
From a strategic point of view, this is a master stroke
Elon positions Grok as the primary AI assistant while harvesting massive datasets for model improvement.
The move follows Elon's pattern of catching up through unconventional speed—building Colossus compute clusters in 8 weeks versus the industry standard of 2 years. Combined with Microsoft partnerships announced at Build, Grok is rapidly becoming a serious AI platform competitor despite starting years behind.
Why does it matter?
Organizations that secure large user bases and high-quality training data will maintain competitive advantages even if their current technology lags. Strategic partnerships and unconventional moves can overcome technical deficits faster than pure R&D investment.
July 10th: The AI Agent Reality Check

“AI agents are coming. But do you know what you’re doing?” July 10th’s Gracious AI Agents Masterclass cuts through endless AI hype to tackle what business leaders actually need: building autonomous AI that carries your voice, values, and vision when you’re not there.
12 spots. 4 hours. One comprehensive framework for deploying AI agents the right way, not just the fast way. The curriculum addresses real barriers—reliability doubts, alignment concerns, ethics questions—that separate successful AI implementation from expensive failures. “Everything you need to begin smart, not just fast.”
Why does it matter?
The gap between AI capability and organizational readiness is widening rapidly. Companies that deploy thoughtful AI agents will gain sustainable competitive advantages, while those rushing into automation without proper frameworks will face reliability issues and stakeholder resistance. The 12-spot limit reflects the intensive preparation required for quality AI implementation rather than superficial tool training.
The Developer Displacement Signal

via Voronoi
Software developer employment in the US has dropped 20% since 2018, with AI acceleration driving the decline. Even when AI can't fully replace developers, one skilled programmer with AI tools can accomplish what previously required 100-person teams. The trend extends beyond tech. McKinsey quietly reduced workforce by 10% using internal AI called "Lilly."
Consulting firms that previously measured success by billable hours are shifting to AI-augmented delivery models. The job market is already reshaping industries while flying under mainstream media radar.
Why does it matter?
The job displacement isn't theoretical or future-oriented—it's happening now across high-skill professions. Organizations that help workers transition to AI-augmented roles will retain talent and capability, while those ignoring the shift will lose their most valuable contributors to more forward-thinking competitors. The window for proactive workforce transformation is closing rapidly.
June 11th: Signals Versus Noise Masterclass
"What most AI events offer is 'we're going to show you how to use ChatGPT to summarize emails,'" Aragorn explains the limitation of current AI education. "That's not what we're trying to do. Those waves of abundance are arriving 10-15 years in the future. But to get from today to there, we have to go through a whole decade of massive change."
The masterclass combines strategic keynotes with practical workshops designed to translate insights into actionable plans. Rather than tool tutorials, the focus is identifying signals that separate genuine transformation from hype, preparing leaders for the decade of unprecedented change ahead.
![]() | That’s all for this week 🫢 Want your team to surf the waves of change? Send us an email |