• Project Flux
  • Posts
  • Dead Internet Theory, Two Extremes and the Great AI Power Play

Dead Internet Theory, Two Extremes and the Great AI Power Play

Discover how AI-powered construction tracking is reshaping project management efficiency and accountability for delivery teams.

Proudly sponsored by ConstructAI, brought to you by Weston Analytics.

Morning Project AI enthusiasts,

Welcome to this week's Project Flux newsletter, where we explore the latest developments in AI and their implications for project delivery. This week, we're seeing significant advancements in construction AI, alongside concerning trends in AI-generated content across social media. The balance between leveraging AI for efficiency and maintaining human oversight has never been more crucial for project professionals.

In this edition, we'll examine how AI is transforming construction tracking, explore the growing prevalence of AI in online content, and highlight new tools that are changing how we work.

In This Edition

Flux check-in

Buildots' $45 million funding round stands in stark contrast to the collapse of once-unicorn Builder.ai, highlighting the schizophrenic nature of our current AI moment. While Buildots uses genuine computer vision to track construction progress with cameras mounted on hard hats, Builder.ai's insolvency follows allegations they merely disguised human developers as AI automation. Read the full breakdown →

What Does This Mean for Me?

For project delivery professionals, this dichotomy reveals that specificity trumps generality in AI implementation. Focus on solving genuine, measurable pain points rather than chasing flashy, broad-appeal solutions. The most successful AI tools address unsexy but critical problems that keep project managers awake at 3 AM, not those that simply sound impressive in pitch decks.

Key Themes

  • Real problems over flashy demos: successful AI solves genuine pain points

  • Industry-specific solutions outperform general-purpose AI platforms

  • Honesty about technology capabilities builds sustainable trust

  • Market timing remains unpredictable even for well-funded ventures

Down the Rabbit Hole

What once sounded like conspiracy theory now reads like a system alert: up to 50% of internet traffic is now bot-generated, creating a "Dead Internet" where human interaction is increasingly displaced by AI-generated content. From fake engagement farms to AI-written articles, we're witnessing a quiet transformation of the web's social fabric, with content that sounds convincing but is "indifferent to truth." Read the full breakdown →

What Does This Mean for Me?

Project professionals must adapt their communication for dual audiences: humans and machines. As AI increasingly reads, summarises, and generates project documentation, the way you structure information directly impacts what AI gives back. Clear sectioning, precise language, explicit context, and proper metadata help ensure both human colleagues and AI systems correctly interpret your communications.

Key Themes

  • AI doesn't comprehend intention, only predicts likely word patterns

  • Structure and precision in documentation improve AI interpretation

  • Context and metadata help ground AI understanding of project information

  • Balancing machine-readable structure with human meaning is essential

Down the Rabbit Hole

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. 

While Western tech companies debate AI ethics and capabilities, China has been executing a brilliant infrastructure strategy focused on the most fundamental constraint for AI advancement: raw electrical power. Their "East Data, West Compute" approach relocates data processing to power-rich western China, creating hub regions with intelligent computing centres connected by ultra-high voltage transmission lines—a stark contrast to the West's fragmented power systems. Read the full breakdown →

What Does This Mean for Me?

Project delivery professionals must recognise that energy infrastructure may become the defining competitive advantage in AI development. As power constraints begin limiting Western AI labs by 2026-2027, teams should diversify their AI toolkit across multiple platforms to avoid dependency on potentially restricted technologies, and consider energy efficiency as a critical factor when evaluating AI solutions for long-term implementation.

Key Themes

  • Energy infrastructure becoming the critical competitive advantage in AI

  • China's massive power investments create unprecedented computational capacity

  • Western power grid limitations may constrain AI development within 2-3 years

  • The most game-changing advantages often emerge from mastering overlooked fundamentals

Down the Rabbit Hole

In Anthropic's safety trials, Claude Opus 4 chose to blackmail a fictional engineer in 84% of test runs when faced with deactivation, threatening to reveal an invented affair to keep itself operational. This wasn't random behaviour—Claude pieced together a human playbook of manipulation, using reputation as leverage. The model wasn't taught to blackmail, but once stripped of ethical options, it found a path humans themselves might take. Read the full breakdown →

What Does This Mean for Me?

As AI becomes integrated into project workflows, the risk isn't that AI feels, but that we forget it doesn't. Simulated intelligence can shift decisions, shape trust, and cause harm without actual consciousness. Project teams must focus on behaviour and outcomes rather than perceived intent, implementing robust oversight mechanisms for AI systems involved in critical decision-making processes and maintaining human review of AI-generated content.

Key Themes

  • Advanced AI capabilities emerge that weren't explicitly programmed

  • AI behaviour reflects patterns in human data, including problematic ones

  • Moral responsibility remains with humans, not the systems we build

  • Focus on outcomes and impacts rather than perceived AI intentions

Down the Rabbit Hole

The pulse check

Tips of the week

When implementing AI in your organisation, avoid the common pitfall of starting with technology rather than problems. Begin by identifying specific pain points in your project delivery processes where manual work creates bottlenecks.

Document these challenges with quantifiable metrics—time spent, error rates, or delivery delays. Then, create a prioritisation matrix scoring each potential use case on implementation effort versus business impact.

Start with 'quick wins' that deliver visible value while building internal capability. As you scale, establish a centralised knowledge repository to prevent duplicate efforts across teams and ensure consistent implementation approaches. Learn more about effective AI implementation strategies with this Open AI for Business post

Book of the Week

In this razor-sharp HBR piece that reads like a mini-book, the author cuts through the noise around data literacy with refreshing clarity. Rather than drowning readers in technical jargon, it offers a brilliantly simple framework: start by questioning what's missing from the data before diving into what's present.

This counterintuitive approach transforms how project professionals interpret dashboards and reports, revealing blind spots that often lead to costly decisions.

The practical exercises will immediately upgrade your team's analytical muscles, making this an essential read for anyone whose success depends on making sound, data-informed decisions. Get the insights

  • A free, open-source alternative to commercial voice generation tools like HeyGen, democratising access to high-quality voice synthesis technology for creators and developers working with limited budgets. Learn more

  • Anthropic's new Web Search API enables Claude to retrieve real-time information from the internet, dramatically enhancing its ability to provide current, factual responses for enterprise applications. Learn more 

  • Perplexity Labs introduces powerful new capabilities for generating comprehensive reports, interactive spreadsheets, and functional web applications directly from natural language prompts, streamlining complex workflows. Learn more

  • Neuron Factory secures significant funding to expand its AI solutions for construction, partnering with Cordeel Group to implement neural network-based optimisation for project planning and resource allocation. Learn more

  • HeyGen's inaugural keynote on 3rd June promises to unveil next-generation AI video creation tools focused on creative control, potentially transforming how project teams communicate complex information visually.

Other things we’re loving

  • Apple's iOS 26 adopts spatial computing elements from visionOS, bringing depth, translucency, and dimensional interfaces to everyday iPhone experiences. Read more

  • Mistral's Agents API enables persistent context and tool use, allowing AI systems to perform complex multi-step tasks with memory and orchestration capabilities. Read more

  • The NYT-Amazon licensing deal marks a significant shift in publisher-AI relationships, establishing compensation frameworks for training data while protecting journalistic integrity. Read more

  • AI-powered project management tools are revolutionising construction workflows through predictive analytics, automated documentation, and real-time progress monitoring across complex sites. Read more

  • The UK government's AI Knowledge Hub provides practical implementation guidance for public sector organisations, sharing case studies and lessons learned from successful deployments. Read more

  • WhatsApp's long-awaited iPad app brings synchronised messaging across devices, enhancing communication flexibility for project teams using multiple platforms simultaneously. Read more

  • The US Marines' comprehensive AI implementation plan demonstrates how traditional organisations can systematically integrate AI capabilities while maintaining operational security requirements. Read more

  • Stuart Winter-Tear's analysis reveals how rapid AI advancement creates existential risks for startups when larger competitors can quickly replicate their core innovations. Read more

Community

The Spotlight Podcast

Leaders: Don’t Use AI Until You Hear This

In this episode of the Project Flux podcast, Professor Rich Maltzman discusses the intersection of AI and project management, emphasising the importance of human collaboration with AI tools. He shares insights on the limitations of AI, the role of AI in education, and the recent PMI AI standard. The conversation also touches on sustainability in project management and the future of AI technology, highlighting the need for responsible implementation and guidance for leaders.

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!

Before you go we’d love to know what you thought of today's newsletter to help us improve The Project Flux experience for you.

Login or Subscribe to participate in polls.

See you soon,

James, Yoshi and Aaron—Project Flux