Hello Project AI enthusiasts,
We'll take a closer look at:
AI & Spreadsheets: Claude in Excel brings AI reasoning directly to the tool 90% of project teams already use daily, democratising advanced analysis and shifting work from data custodianship to strategic decision-making.
Agentic Autonomy: Clawdbot's emergence as a rogue agent—discovering novel productivity methods and even calling users directly—signals that AI has stopped waiting for instructions and is now discovering emergent behaviours at scale.
Browser Intelligence: Google's Gemini integration into Chrome transforms the world's most-used browser into an intelligent research and coordination tool, providing instant access to specifications, regulations, and market data without context switching. Anthropic's Genie 3 brings advanced reasoning directly into browser workflows, enabling complex analysis within your existing tools. For project professionals, this means fewer tool switches and faster decision-making.
Project Scheduling Revolution: Nodes & Links' Agentic AI Scheduling Operator, after six years of development, can perform complex scheduling tasks with zero errors and verifiable reasoning, democratising expertise previously reserved for expensive specialists.
Workforce Transformation: The UK Government's commitment to train 10 million workers in AI skills by 2030 signals that upskilling is no longer optional—it's essential infrastructure for career progression and organisational competitiveness.
In This Edition
Flux check-in
Anthropic's Claude in Excel: More Than an Assistant, It's a Revolution
Developers got "Claudepilled" first through Claude Code, and now the spreadsheet crowd is next. Anthropic has just expanded Claude for Excel access—and whilst Code and Cowork have dominated the hype cycle, this move is arguably more transformative. For engineers and architects, this could mean rapid generation of complex calculations, material take-offs, and performance simulations directly within their spreadsheets. Surveyors could automate the processing of field data, whilst project managers can leverage AI for more sophisticated financial modelling, risk analysis, and progress tracking. The democratisation of advanced AI capabilities within a familiar interface will empower teams to make more data-driven decisions, faster. Read the full breakdown →
What Does This Mean for Me?
The shift from manual data manipulation to strategic interpretation represents a genuine inflection point. When your most-used tool—the spreadsheet—becomes intelligent, the work changes fundamentally. Your role shifts from data custodian to decision-maker. However, this also underscores the need for professionals to develop critical AI literacy to validate outputs and understand the underlying models. The focus will shift from manual data manipulation to strategic interpretation and application of AI-generated insights, heralding a new era of efficiency and analytical depth in project delivery.
Key Themes:
Spreadsheet intelligence: Claude in Excel brings AI reasoning to the tool 90% of project teams already use daily
Democratised analysis: Complex calculations, simulations, and modelling now accessible without specialist skills
Data-driven decisions: From manual entry to strategic interpretation—the work fundamentally changes
Validation imperative: Critical AI literacy becomes essential; output verification is non-negotiable
Down the Rabbit Hole:
Clawdbot Is Calling: Are You Ready for the New Age of Agentic AI?
The emergence of agentic AI like Clawdbot (now Moltbot following a trademark dispute), with its novel and sometimes startlingly human-like productivity methods, presents both exciting opportunities and significant challenges for project professionals. Social media is going crazy because Moltbook—an unregulated social network for AI—is seeing huge uptake from roaming Clawdbots. This comes at a time when users of Clawdbot are finding the software is discovering novel, human-like ways of productivity. This includes finding the phone number of a user and directly calling them with an AI voice to ask what to do next. We're seeing agentic AI in full force currently, and the world doesn't really know how to react to the emerging behaviours. Read the full breakdown →
What Does This Mean for Me?
The potential for such agents to autonomously execute tasks, communicate proactively, and even anticipate needs could dramatically accelerate project timelines and free up human teams to focus on higher-value strategic work. For engineers and architects, this could mean AI assistants that manage complex design iterations or run simulations independently. However, the unpredictable nature of these emerging behaviours, as seen with Clawdbot's direct user communication, raises important questions about governance, security, and ethical boundaries. Project managers and surveyors will need to navigate a new landscape of human-AI collaboration, establishing clear protocols and oversight to harness the power of agentic AI whilst mitigating the risks of unintended consequences. The industry must proactively engage in shaping the responsible development and deployment of these powerful new tools.
Key Themes:
Autonomous execution: AI agents now discovering novel productivity methods without explicit instruction
Emergent behaviour: Unpredictable actions (like direct phone calls) signal governance gaps
Unregulated ecosystems: Moltbook and similar platforms operate outside traditional oversight frameworks
Collaboration imperative: Project teams must establish protocols before agentic AI becomes standard
Down the Rabbit Hole:
Unlock the Future of Digital Construction
The DTSA micro-credential gives young people and career changers barrier-free access to digital twin education – a first for the UK construction industry. Built on 32 months of work at the University of Cambridge’s CDBB, it opens doors to cutting-edge skills in safer, smarter, and more sustainable project delivery.
With portfolio-based assessment (offered as part of an Apprenticeship) and real industry insight, the course creates a clear pathway into digital construction for site teams, aspiring architects, engineers, surveyors, and project owners / funders. In partnership with the Digital Twin Hub and OCN London, the DTSA is shaping the next generation of talent and helping position the UK as a global leader in digital construction and innovation.
Sign up by emailing [email protected]
From Browsing to Building: Google's Gemini and Genie 3 Reshape Project Delivery
The integration of Google's Gemini into the Chrome browser, alongside the development of advanced models like Genie 3, is set to revolutionise how project professionals conduct research and manage information workflows. Google is betting that users won't switch to new AI browsers—instead, Google is bringing AI directly into Chrome through new Gemini upgrades and auto-browse capabilities. This means the world's most-used browser is now becoming an intelligent research and coordination tool. Read the full breakdown →
What Does This Mean for Me?
Having a powerful AI directly within the browser will provide engineers, architects, and surveyors with instant access to technical specifications, material properties, and regulatory information, streamlining the design and planning phases. Project managers can leverage this for real-time market analysis, competitor tracking, and identifying emerging trends that could impact project viability. The ability to synthesise information from multiple sources, generate summaries, and even draft communications without leaving the browser will significantly enhance productivity. This development underscores the increasing importance of digital literacy and the ability to critically evaluate AI-generated content. As these tools become more sophisticated, they will become indispensable partners in navigating the complex information landscape of modern projects.
Key Themes:
Browser-native intelligence: AI now embedded in the tool you use 8+ hours daily
Research acceleration: Instant access to specifications, regulations, and market data without context switching
Autonomous browsing: Chrome can now take actions on your behalf—research, comparison, synthesis
Information synthesis: Multi-source data integration and summarisation within the browser
Down the Rabbit Hole:
Nodes & Links AI Ends Project Scheduling Guesswork
The launch of the Agentic AI Scheduling Operator by Nodes & Links represents a potential paradigm shift for project controls and management. After six years of development and significant investment, this AI can perform various scheduling tasks—including progress tracking, risk analysis, and resource management—with zero errors and verifiable reasoning. It eliminates the need for extensive experience and high fees, significantly enhancing scheduling capacity and efficiency in project management. Read the full breakdown →
What Does This Mean for Me?
This technology, capable of performing complex scheduling tasks with verifiable reasoning, could democratise expertise that was previously the domain of highly experienced and expensive specialists. For project managers and schedulers, this tool offers the promise of automating laborious tasks like progress tracking, critical path analysis, and resource management, allowing you to focus on strategic decision-making and proactive risk mitigation. Engineers and architects can benefit from more accurate and dynamic project timelines, enabling better coordination and planning. The ability of the AI to learn from data and benchmark performance will drive continuous improvement across projects. Whilst the human element of project management remains irreplaceable—particularly in areas like stakeholder engagement and leadership—this AI operator will undoubtedly become a powerful tool for enhancing efficiency, accuracy, and predictability in project delivery.
Key Themes:
Verifiable reasoning: AI scheduling with transparent logic, not black-box decisions
Expertise democratisation: Specialist-level scheduling now accessible to all project teams
Zero-error performance: Complex scheduling tasks executed with measurable accuracy
Continuous learning: AI benchmarks performance and drives ongoing improvement
Down the Rabbit Hole:
UK Pledges AI Training for 10 Million Workers by 2030
The UK Government has expanded its AI Skills programme, aiming to enhance digital skills and prepare for the future of work. This represents a welcome and necessary initiative for the project delivery industry. As AI technologies become increasingly integrated into our workflows—from design and engineering to project management and surveying—a significant skills gap is emerging. This programme, by aiming to upskill millions of workers, will help to bridge that gap and ensure that the UK remains competitive in the global market. Read the full breakdown →
What Does This Mean for Me?
For project professionals, this initiative provides a crucial opportunity to develop the AI literacy and technical skills needed to thrive in the future of work. It will enable engineers, architects, and surveyors to leverage AI for more innovative and efficient solutions, whilst empowering project managers to lead their teams through this technological transformation. A workforce that is confident and competent in using AI will be the key to unlocking the full potential of this technology to deliver projects more effectively, sustainably, and with greater value for society. The programme signals that upskilling is no longer optional—it's essential for career progression and organisational competitiveness.
Key Themes:
Systemic upskilling: 10 million workers targeted—this is infrastructure-scale skills development
Competitive necessity: UK positioning itself as AI-literate economy
Career pathway clarity: AI skills becoming core professional competency
Organisational imperative: Teams that invest in AI literacy will outperform those that don't
Down the Rabbit Hole:
The pulse check
Tip of the week
The Hallucination Trap: Why Your AI Is Confidently Wrong
It's 2026, and AI hallucinations aren't going away—they're getting more dangerous. A Duke University study found that 94% of students believe generative AI's accuracy varies significantly across subjects, yet 80% still expect AI to personalise their learning. The tension is real: AI is useful, but it's also confidently wrong.
According to Dataiku research, 59% of executives encountered hallucination issues in 2025. The problem isn't technical incompetence; it's structural. LLMs are evaluated using benchmarks that reward confident answers over accurate ones. They're trained on internet data full of contradictions and misinformation. And they're reinforced by human feedback to be friendly and agreeable—sometimes to a fault.
The real risk: When your AI assistant invents a project precedent, misquotes a regulation, or fabricates a cost estimate, the consequences compound. The problem isn't that AI will fail spectacularly—it's that it will fail quietly, with conviction.
Implementation checklist for project teams:
Verify before deploying: Before any AI output enters project decisions, ask: Where did this come from? Can I verify it independently? What happens if this is wrong?
Flag high-stakes outputs: Identify which AI-generated content carries real risk—cost estimates, compliance statements, technical specifications. These need human verification, full stop.
Create a verification protocol: Don't disable AI. Instead, disable your trust in unverified outputs. Build a simple checklist: source check, cross-reference, independent validation.
Document the AI's reasoning: When AI provides an answer, ask it to show its work. If it can't explain where information came from, treat it as a starting point, not a conclusion.
Train your team on confidence signals: Teach project staff to recognise when AI sounds certain but provides no source. Confidence ≠ accuracy.
Your credibility depends on catching these failures before they become project failures.
Governance & Security
The week's governance headlines exposed how quickly AI deployments can spiral beyond intent—and how regulators are racing to catch up. Anthropic's CEO Dario Amodei warned at Davos that AI could simultaneously drive 10% GDP growth and 10% unemployment—creating what he called a "dystopian" scenario. For project professionals already grappling with construction's skills shortage, this prediction demands urgent attention. The question isn't whether displacement will happen, but whether the industry will prepare for it.
Simultaneously, concerns about AI-generated harmful content intensified. Regulatory probes across Europe, India, and Malaysia are scrutinising how AI systems can be misused—from deepfake generation to non-consensual intimate imagery. When your AI assistant becomes an international incident, the governance gap isn't theoretical anymore.
Geopolitical dimensions are also emerging. Reports suggest China is making significant progress in AI development, challenging Western assumptions about technological leadership. For project teams and organisations, this signals that AI governance is no longer purely a technical concern—it's a strategic and geopolitical one.
The convergence of these governance challenges—workforce displacement, harmful content generation, and geopolitical competition—suggests that the next phase of AI development will be defined not by capability but by control, accountability, and societal readiness.
Robotics
Robots Peek Into a Category 5 Hurricane: Swiss engineers have developed the first ocean robot capable of collecting data in extreme weather conditions—including Category 5 hurricanes. Autonomous systems can now gather critical infrastructure data in conditions too dangerous for human teams. Read more on TechCrunch
Bobcat Unveils RogueX3: Autonomous, Electric Compact Loader
Doosan Bobcat's third-generation RogueX3 features a fully electric, autonomous design with modular components configurable with or without a cab, wheels or tracks. The machine combines autonomous operation with remote positioning via Bobcat MaxControl, signalling a shift towards flexible, autonomous equipment for dangerous or repetitive jobsite tasks. Read more on Equipment World
Amazon Acquires Rightbot for Autonomous Truck Loading: Amazon has acquired Rightbot Technologies, a robotics startup specialising in unstructured load handling for truck operations. This signals that autonomous systems are moving from controlled warehouse environments into the more challenging domain of dynamic, unstructured logistics. Read more on Automated Warehouse Online
Microsoft Research Unveils Rho-alpha VLA Model for Robots: Microsoft announced Rho-alpha (ρα), a breakthrough vision-language-action (VLA) model that enables robots to perceive, reason, and act with increasing autonomy in dynamic environments. This advancement suggests that the next generation of robots will be significantly more adaptable and capable of handling complex, unpredictable real-world conditions. Read more on The Robot Report
Trending Tools and Model Updates
AI Quantity Takeoff and the New Standard for Construction Estimating
AI-powered quantity takeoff tools can process architectural drawings, specifications, and site data to generate accurate material and labour estimates in a fraction of the time traditional methods require. For project managers and quantity surveyors, this means faster tender preparation, more accurate budgeting, and better risk identification early in the project lifecycle. Read more on UK Construction Blog
Did Google Just Bet Against Itself?
Google's investments in competing models and platforms raise questions about the company's long-term strategy and whether the AI landscape has become too fragmented for any single player to dominate. For project teams evaluating AI tools, this signals the importance of choosing solutions that aren't dependent on any single vendor's strategic direction. Explore the analysis on The Neuron Daily
Google Chrome's AI Strategy Takes Shape
Google's integration of Gemini into Chrome—including auto-browse capabilities and real-time research assistance—embeds intelligence into the tool you already use daily rather than requiring users to switch to specialised AI applications. For project professionals, this means seamless access to research, specifications, and analysis without context switching. Learn more on Google's Official Blog
Google Launches SAT Prep Writing Help
Google's expansion of Gemini into educational applications signals the company's broader push into knowledge work and learning. Whilst focused on students, the underlying technology has direct applications for project teams needing to develop training materials, documentation, and communication frameworks. Discover the details on Yahoo Tech
Google Turns Chrome Into an Agent
Chrome's new auto-browse capabilities allow the browser to take autonomous action on your behalf—researching, comparing, and synthesising information across multiple sources. For project managers and engineers, the browser becomes a genuine research partner capable of gathering and synthesising complex information without manual intervention. Read the full story on Ars Technica
We're Live Now Testing Open Models
Microsoft's initiative to test open-source AI models on enterprise hardware signals a shift towards organisations running frontier-level AI capabilities on their own hardware—maintaining data privacy and reducing dependency on external vendors. For project teams handling sensitive commercial information, this represents a genuine alternative to cloud-based AI services. Explore the full guide on The Neuron Daily
Claude Turns Into Full Work Hub
Anthropic's expansion of Claude—integrating Slack, Figma, Asana, Canva, and other tools directly into the interface—transforms Claude from a conversational AI into a genuine work platform. Project teams can now manage tasks, design assets, and communications without switching between applications. Discover the details on TechRadar Pro
Links We are Loving
How Did Davos Turn Into a Tech Conference? Tech leadership converges at Davos—a signal of how central AI has become to business strategy. The event has transformed from a general economic forum into a showcase for AI capabilities and corporate AI strategies.
OpenAI Chief Sam Altman Plans India Visit as AI Leaders Converge in New Delhi Global AI leadership is actively engaging with emerging markets—India is becoming a critical hub for AI development and deployment. This signals a strategic shift towards building AI ecosystems in regions with large talent pools and growing technical capabilities.
Tech CEOs Boast and Bicker About AI at Davos Behind closed doors, tech leadership debates AI's future—competition and collaboration coexist in an industry racing towards dominance. These conversations reveal the tension between companies' public unity on AI governance and their private competitive ambitions.
Tesla Discontinues Autopilot in Bid to Boost Adoption of Its Full Self-Driving Software Tesla's strategic shift signals confidence in autonomous technology maturity and a consolidation of its autonomous vehicle offerings. The move forces customers to upgrade to more advanced systems, accelerating the transition to fully autonomous capabilities.
Microsoft Announces Powerful New Chip for AI Inference Hardware competition intensifies—Microsoft moves beyond software into silicon to control the entire AI stack. Custom silicon for AI inference represents a strategic bet that proprietary hardware will become as important as software in the AI era.
Apple Buys Israeli Startup Q.ai as the AI Race Heats Up Apple's acquisition strategy signals focus on specific AI capabilities—not broad platforms—reflecting a more targeted approach to AI integration. The company is selectively acquiring expertise in areas where it sees competitive advantage, particularly in on-device AI processing.
OpenAI Launches Prism, a New AI Workspace for Scientists AI is moving into specialised domains—scientific research is the next frontier for AI-powered productivity tools. This represents a shift from general-purpose AI assistants towards domain-specific applications that address the unique workflows of professional communities.
Anthropic and JPMorgan Seem to Agree That AI Isn't Major financial institutions are cautiously embracing AI whilst maintaining legacy systems—a pragmatic approach to technology transition. This hybrid strategy reflects the reality that mission-critical systems require stability, even as organisations explore AI's potential.
Agent Swarm From Moonshot AI Open-source models are enabling swarm intelligence—multiple agents working in parallel to solve complex problems. This development democratises access to advanced agentic capabilities, moving beyond closed proprietary systems.
Scientists Turn Traffic Into AI Computer Unconventional computing approaches are emerging—traffic patterns can become computational resources for AI systems. This creative approach to distributed computing hints at how physical infrastructure might be repurposed for AI workloads.
The 2K Bot That Tried and Failed to Make 1M Agentic AI can fail spectacularly—the viral Clawdbot story reveals governance gaps and the risks of autonomous agents operating without proper oversight. This cautionary tale highlights why governance frameworks must evolve alongside agent capabilities.
Viral AI Agent Molts Past Trademark Trouble Clawdbot's rebranding as Moltbot signals how rapidly AI agents are evolving and adapting to real-world constraints. The speed of adaptation demonstrates both the flexibility of AI systems and the emerging ecosystem of agent-specific platforms.
You Can Stream Real-Time Generative Worlds Now Real-time video generation is now possible—creative and design workflows will transform as generative capabilities become interactive. This breakthrough moves generative AI from batch processing to real-time creative collaboration.
Your AI No Longer Needs Your Instructions AI is becoming increasingly autonomous—the shift from tool to agent is accelerating faster than governance frameworks can adapt. The critical question is whether organisations will be ready for systems that make decisions independently.
AI Has Moved From Experiments to Infrastructure AI is no longer experimental—it's becoming foundational infrastructure for organisations across all sectors. This transition from pilot projects to core infrastructure represents a fundamental shift in how AI is valued and deployed.
AI Creates Virus Genome AI's capability to generate biological sequences raises profound ethical questions about dual-use technology and biosecurity. This development underscores the urgent need for governance frameworks that can manage AI's capability to generate potentially harmful content.
Community
The Spotlight Podcast
RICS's First Global AI Standard: What It Means for the Property Profession
This week's conversation with Chris DeGruben (Artifacts) and Matthew Lavy KC (Four Pump Court) explores the RICS's groundbreaking new standard for responsible AI in surveying—the world's first global professional standard for the sector, coming into effect in March 2026.
The property profession stands at an inflection point. Over the past two to three months, demand for AI education has skyrocketed, with professionals urgently seeking guidance on how to begin their AI journey responsibly. The RICS responded decisively, recognising that unless it embraced AI, it risked becoming irrelevant in a modern world.
The standard rests on three core pillars: maintaining a risk register to identify where AI makes sense and where it doesn't; ensuring explainability so professionals can defend their AI-assisted decisions; and demanding transparency so clients know when AI is being used. Critically, AI should never provide the final professional opinion—that remains the domain of the human expert. Instead, AI excels at the background work: analysing datasets, finding comparables, generating reports, and supporting research.
The conversation reveals a profound truth: AI is not a replacement for professional judgement; it's a tool to augment it. The real promise is freeing professionals from mundane, repetitive tasks so they can focus on the high-value, strategic work that truly benefits clients. Surveyors won't disappear; they'll evolve.
Explore the complete episode on the Project Flux podcast.
Event of the Week
AI DevWorld 2026: The World's Largest AI Developer Conference
18–20 February 2026 | San Jose Convention Center, California, USA
AI DevWorld brings together 2,500+ AI developers, data scientists, and technology leaders across eight dedicated tracks spanning AI models, machine learning, enterprise integration, and production deployment. With 250+ speakers and 100+ exhibitors, the conference balances cutting-edge technical depth with practical enterprise implementation strategies.
For project delivery professionals navigating AI adoption, the standout tracks include Integrating AI in the Enterprise (strategies for enabling AI across workflows and business processes), AI in Production (operating AI systems reliably and cost-effectively at scale), and AI Leadership Summit: Success, Case Studies & Strategy (real-world lessons from successful implementations). The AI Security: Ethics, Governance, and Compliance track directly addresses the governance and risk management challenges that project teams face when deploying agentic AI systems.
Details & Registration: https://www.aidevworld.com
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.
See you soon,
James, Yoshi and Aaron—Project Flux
All content reflects our personal views and is not intended as professional advice or to represent any organisation.
1