Morning Project AI enthusiasts,

This week has been a masterclass in how quickly the AI world can pivot. From political fallouts reshaping funding priorities to Apple's sobering research on reasoning limits, we're witnessing the growing pains of an industry still finding its footing.

Yet beneath the turbulence, there's something more hopeful emerging: a clearer picture of what AI fluency actually means, and why human creativity remains irreplaceable.

In this edition, we explore the Musk-Trump rift's ripple effects, dive into Apple's latest reasoning research, examine the surprising job market data, and discover what true AI literacy looks like in 2025 and highlight new tools that are changing how we work.

In This Edition

• The Musk-Trump Rift: When Politics Reshapes Innovation 

• Apple's Reality Check: The Illusion of AI Reasoning 

• The AI Jobs Paradox: Why Doomsday and Prosperity Predictions Are Both Right 

• Defining AI Fluency: What It Really Means to Be Literate in 2025 

• Why We Buy the Story, Not the Art: Human Creativity in the AI Age

Flux check-in

The Musk-Trump Rift: When Politics Reshapes Innovation

The fragile alliance between two of America's most influential figures has cracked, and the reverberations are reshaping AI's political landscape. When Trump threatened to cancel federal contracts with Musk's companies, Tesla lost over £120 billion in a single day—its largest ever session drop. Read the full breakdown →

What Does This Mean for Me?
Political volatility directly impacts AI funding and adoption incentives. When public money shifts from smart infrastructure to legacy sectors, the business case for intelligent systems weakens—not because they don't deliver, but because the ecosystem no longer rewards them.

Key Themes: 

• Innovation built on individual influence creates systemic risk
• Political moments redirect technological progress
• AI adoption follows where incentives lead
• Competing ecosystems may emerge from this divide

Down the Rabbit Hole:
• ChatGPT conversation on the fallout

Apple's Reality Check: The Illusion of AI Reasoning

Apple's latest research paper takes aim at a fundamental question: do large reasoning models actually reason, or are we projecting logic onto sophisticated pattern-matching? Their controlled study of logic puzzles reveals something striking—as problems grow complex, models don't just perform worse, they often stop trying altogether. Read the full breakdown →

What Does This Mean for Me?
Understanding AI's reasoning boundaries helps set realistic expectations for project delivery. These models excel at surface-level planning but struggle with deep, recursive thinking—knowledge that's crucial for scoping AI implementations effectively.

Key Themes: 
• Models exhibit "learned frugality" when sensing complexity limits
• Reasoning exists but remains shallow and heuristic-driven
• Spatial simulation challenges expose text-based limitations
• Strategic retreat may indicate computational maturity

Down the Rabbit Hole:
• Apple's original research paper

Together with Cogram

Power your construction bids with AI

Cogram’s AI-assisted RFP bidding tool writes tailored RFP proposals in minutes instead of weeks.

• Automatically extract key details from the RFP — including scope, submission requirements, deadlines, and evaluation criteria — to easily make a go/no-go decision.

• Cogram’s AI will then reference your firm’s knowledge base and past proposals to draft tailored proposals within minutes.

• Use AI-assisted editing tools to review, cross-check data, and make improvements remarkably fast. 

_{Try Cogram on a real RFP in a 20-minute live meeting.}

The AI Jobs Paradox: Why Doomsday and Prosperity Predictions Are Both Right

A PwC study analysing nearly a billion job adverts reveals a surprising truth: AI isn't just displacing workers—it's making some more valuable. Those with AI skills like prompt engineering command a 56% wage premium, while employer demand for formal degrees declines faster in AI-exposed roles. Read the full breakdown →

What Does This Mean for Me?
The job market is bifurcating. AI augments rather than replaces many roles, but success requires active skill development. Practical AI capabilities are becoming more valuable than traditional qualifications in many sectors.

Key Themes: 
• AI creates wage premiums for skilled practitioners
• Job growth accompanies AI adoption in many sectors
• Practical capabilities trump formal qualifications
• Role upgrading occurs more than wholesale replacement

Down the Rabbit Hole:
• UK Government's Microsoft 365 Copilot findings
• Axios report on white-collar AI impact

Defining AI Fluency: What It Really Means to Be Literate in 2025

Zapier's recent decision to require AI fluency for all new hires signals a workplace transformation. But what does "AI-fluent" actually mean? This comprehensive framework breaks down 14 core competencies, from hands-on tool mastery to ethical judgement and collaborative augmentation. Read the full breakdown →

What Does This Mean for Me?
AI fluency isn't just about using tools—it's about thinking with machines. This framework provides a roadmap for developing the skills that will define professional competence in an AI-first world.

Key Themes: 
• AI fluency encompasses technical, ethical, and collaborative skills
• Continuous learning habits matter more than static knowledge
• Domain application requires tailored approaches
• Cultural fit involves treating AI as the default work interface

Down the Rabbit Hole: 

• Wade Foster's thoughts on AI literacy
• Robert Scoble's perspective on AI fluency
• Additional Scoble insights

Why We Buy the Story, Not the Art: Human Creativity in the AI Age

As AI-generated art floods the market, a fundamental truth emerges: we don't buy art for its technical execution—we buy it for the human story behind it. This shift reveals why human creativity remains irreplaceable, even as machines master technique. Read the full breakdown →

What Does This Mean for Me?
In a world where AI can replicate technical skills, the value lies in human narrative, context, and meaning. This principle extends beyond art to any creative or strategic work—the story behind the solution matters as much as the solution itself.

Key Themes:
• Human narrative creates irreplaceable value
• Technical execution becomes commoditised
• Authenticity and context drive purchasing decisions
• Creative work requires human meaning-making

The pulse check

Tips of the week

AI Agents for Beginners

Microsoft released a comprehensive, free course covering agent design, tool use, planning, and multi-agent systems. The hour-long programme defines AI agents as intelligent systems with a brain (LLM), memory, and tools—essential knowledge as agentic AI becomes mainstream.knowledge repository to prevent duplicate efforts across teams and ensure consistent implementation approaches.

Governance & Security

Anthropic is reshaping its approach to sensitive applications, appointing a national security expert to its governing trust whilst unveiling custom AI models for US national security customers.

Meanwhile, the UK government published guidance on human-centred AI scaling and risk management, emphasising responsible deployment practices. These moves signal a maturing approach to AI governance, balancing innovation with security considerations.

Trending tools

• AI Productivity Powerhouses
  ChatGPT's o3 model now tackles tax preparation, legal questions, and document conversion, while new "Record mode" captures meeting notes. Perplexity Labs offers Pro users comprehensive research reports and dashboards. OpenAI's latest business integrations

• Voice Generation Revolution
  Bland's new TTS model clones voices from short snippets, ElevenLabs launched Eleven v3 with 70+ language support, and Hume's EVI 3 creates emotionally aware voices. The line between AI and human speech continues to blur.

• Visual Truth Crisis
  Flux1 Kontext, Mirage Studio, and HeyGen AI Studio make image and video editing so precise that trusting online visuals now depends entirely on source credibility. We're entering the "end of truth" era for visual content.

• Platform Updates

  • Luma improves AI video generation,

  • ChatGPT expands business connections,

  • Manus adds video generation, and

  • Claude Projects supports more materials.

Other things we’re loving

• FDA approves AI breast cancer prediction platform shifting healthcare towards proactive prevention 

• Perplexity processed 780 million queries last month, highlighting AI search adoption 

• Reddit sues Anthropic over training data usage, escalating IP battles 

• Meta invests in nuclear power for AI infrastructure 

• Google quietly releases local AI model app for Android devices 

• Anthropic's AI writes its own blog content with human oversight 

• Remote surgery advances demonstrate AI-enabled precision from miles away 

• AI and BIM in construction risk management sets new efficiency standards 

• UK military AI and robotics developments for battle readiness

• Smart money shifts focus beyond AI models to infrastructure 

• How one teenager leverages AI for productivity and learning 

• Anthropic's thought tracing capabilities reveal model reasoning processes
  

Community

The Spotlight Podcast

The Real Power of Microsoft's AI
A deep dive into Microsoft's AI strategy reveals how the tech giant is positioning itself for the next wave of enterprise adoption. Listen here for insights on hybrid architectures and competitive positioning.

Listen now or read the accompanying blog.

Simon Sinek: You're Being Lied To About AI's Real Purpose
Simon Sinek challenges conventional AI narratives, arguing we're teaching our children not to be human. A provocative discussion on AI's true purpose and its impact on human development.

Events

The PDA Task Force and Gleeds are supporting EY's upcoming AI in Infrastructure event presents a practical implementation framework for government departments and industry partners.

Attendees will gain access to real-world roadmaps for translating theoretical AI benefits into tangible project delivery outcomes, with specific focus on overcoming data quality challenges that often derail digital transformation initiatives. Register now

One more thing

That’s it for today!

See you soon,

James, Yoshi and Aaron—Project Flux