• Project Flux
  • Posts
  • OpenAI's Study Mode Revolution: Transforming Project Team Learning Forever

OpenAI's Study Mode Revolution: Transforming Project Team Learning Forever

From educational AI breakthroughs to Microsoft's record quarter—discover how artificial intelligence is reshaping professional development and project delivery success.

AuthorAuthor

Yoshi Soornack & James Garner
August 05, 2025

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

Morning Project AI enthusiasts,

From OpenAI's quiet launch of Study Mode that could revolutionise education to Microsoft's record-breaking £76.4 billion quarter, we're witnessing fundamental shifts in how artificial intelligence integrates into our professional lives.

Meanwhile, Meta's bold vision for personal superintelligence raises profound questions about the future of human-AI collaboration, and the ongoing Microsoft-OpenAI negotiations highlight the complex challenges of defining what "superhuman AI" actually means.

In this edition: five game-changing stories that every project delivery professional needs to understand.

In This Edition

Flux check-in

OpenAI has quietly launched Study Mode for ChatGPT, and whilst the tech world does its usual song and dance, this could be the most significant development in educational technology we've seen in decades. Unlike traditional AI that spoon-feeds answers, Study Mode asks questions back, wanting to know your skill level, goals, and learning objectives. It's personalised education finally delivered through AI magic—if anyone actually discovers it exists.

What Does This Mean for Me? For project delivery professionals, this represents a fundamental shift in how teams can upskill and adapt to new technologies, methodologies, and industry standards.

Key Themes: • The Great AI Divide: 99% use AI daily, 64% don't know it • Educational apartheid between AI-aware and AI-oblivious learners • Cognitive enhancement versus digital dependency risks • Institutional resistance versus student reality

A landmark Microsoft Research study analysing 200,000 real user conversations reveals the first clear, data-driven picture of how AI is transforming project delivery. The findings are striking: AI isn't replacing project managers—it's eliminating their administrative burden, freeing them for high-value work that only humans can do.

What Does This Mean for Me? Your most skilled professionals can now focus on on-site problem-solving, complex negotiations, and critical stakeholder management whilst AI handles the paperwork.

Key Themes: • Tale of two workforces: augmented office versus unaffected job site • AI as research assistant, not replacement manager • 40% of work conversations show AI supporting, not replacing • Historical precedent: ATM model of job transformation

Down the Rabbit Hole: • Microsoft Research Study • Fortune Analysis

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. 

Microsoft just delivered its strongest quarterly performance in company history, and the numbers tell a story that goes far beyond impressive financial results. With £76.4 billion in revenue, the company has demonstrated that AI and cloud computing aren't just technological trends—they're fundamental business necessities.

What Does This Mean for Me? The productivity gains are measurable: 60%+ time reductions across all job categories, with technical roles seeing up to 76% efficiency improvements.

Key Themes: • £281.7 billion annual revenue with 18% year-over-year growth • Azure surpasses £75 billion revenue, growing 34% • Workplace AI usage jumped from 30% to 43% in three months • Enterprise adoption accelerating beyond experimentation

Mark Zuckerberg has outlined Meta's plan to build "personal super intelligence for everyone in the world"—an AI that knows you deeply and acts as a constant companion through wearable devices. This pivot raises critical questions about technological safety, commercial imperatives, and the future of open-source AI development.

What Does This Mean for Me? The shift from open-source to proprietary AI models could fundamentally change how project teams access and implement AI tools.

Key Themes: • Rhetorical shift from AGI to ASI (Artificial Superintelligence) • Wearable devices as primary interface for personal AI • Commercial pressure driving closed development models • Geopolitical implications of AI openness versus control

Microsoft and OpenAI are negotiating the future of artificial intelligence based on an AGI milestone that's about as clearly defined as "when pigs fly." The stakes? Microsoft's entire AI strategy and £13.75 billion in investment, all hinging on achieving something nobody can properly define.

What Does This Mean for Me? Understanding these definitional challenges is crucial for project leaders making long-term AI investment decisions.

Key Themes: • AGI definition varies from pattern recognition to consciousness • Microsoft seeking 30%+ equity stake in restructured OpenAI • Negotiations based on undefined technological milestone • Project management parallels: defining "done" before starting

The pulse check

Tips of the week

Master Long Context Prompt Optimisation

Anthropic's latest guidance reveals a game-changing technique that can improve your AI response quality by up to 30%. The secret? Place long-form documents at the very top of your prompts—before instructions or questions. This simple repositioning technique is particularly effective when working with multiple complex documents and represents a significant improvement in prompt engineering.

For project delivery professionals juggling multiple technical specifications, client requirements, and regulatory documents, this approach can dramatically improve the quality of AI-generated summaries, analyses, and recommendations. Instead of embedding documents within your questions, lead with the content and follow with your specific requests. The AI's attention mechanism works more effectively when it can process the full context before attempting to respond to your queries.

Robotics

AI Video Companies Pivot to Robotics
AI video generation companies Luma and Runway are pivoting towards robotics and autonomous vehicles, expecting these markets to become major revenue drivers for their "world models."
Read more →

Tesla Optimus Production Updates
Tesla has provided updates on Optimus robot production, with plans to begin limited production for internal use by the end of 2025.
Learn more →

Unitree's £5,900 Humanoid Robot
Unitree has announced a humanoid robot priced at £5,900, making advanced robotics more accessible to consumers and small businesses with advanced mobility and AI integration.
Explore Unitree →

Figure.02 Laundry Capabilities
Figure Robotics has demonstrated their Figure.02 humanoid robot performing laundry tasks, showcasing significant advances in domestic robotics applications.
Watch demonstration →

Governance & Security

The regulatory landscape for artificial intelligence continues to evolve rapidly, with significant developments across multiple jurisdictions this week. Google has announced it will sign the EU's AI Code of Practice, providing guidance on meeting AI Act requirements. This contrasts sharply with Meta's decision not to comply, highlighting growing divisions among tech giants regarding European AI regulation. The voluntary code aims to provide legal certainty for companies navigating the complex regulatory environment.

Meanwhile, the European Union's AI Act has entered a new phase with rules on general-purpose AI models taking effect from 2nd August 2025. This includes requirements for transparency, safety measures, and governance structures. The implementation represents a significant milestone in global AI regulation, establishing precedents that other jurisdictions are likely to follow.

In a surprising development, the Department of Government Efficiency (DOGE) has created an AI-powered "Deregulation Decision Tool" designed to review approximately 200,000 federal regulations and assess which can be eliminated. The tool aims to cut up to 50% of federal regulations within Trump's first year in office, representing a significant use of AI in government reform. This approach demonstrates how AI is being deployed not just for compliance, but for actively reshaping regulatory frameworks themselves.

These developments underscore the complex interplay between technological advancement and regulatory oversight, with different approaches emerging across jurisdictions and organisations.

Toolplay Unified AI Creation Platform
A unified platform for AI creation needs including images, videos, voiceovers, and writing under one account using flexible credits instead of multiple subscriptions. This consolidation approach addresses the fragmentation in AI tools market.
Explore Toolplay →

Shortcut AI for Excel Spreadsheets
An AI agent specifically designed for Excel spreadsheets, automating complex spreadsheet tasks and data analysis. This tool represents the growing trend of AI integration into traditional productivity software.
Try Shortcut AI →

Writer Action Agent for Enterprise
Writer has launched Action Agent, an enterprise AI agent that executes tasks and uses tools in its own environment. The system reportedly outperforms both Manus and OpenAI's Deep Research on benchmarks.
Learn about Writer Action Agent →

Ideogram Character Consistency Model
Character allows users to place a specific person into existing scenes and new outputs from a single reference photo. This tool addresses one of the key challenges in AI-generated content.
Try Ideogram Character →

Runway Aleph Video Editing Tool
Aleph allows users to edit videos by simply describing desired changes. Users can remove cars, swap backgrounds, or change camera angles while maintaining perfect consistency across frames.
Explore Runway Aleph →

Datagrid
Advanced data management and analysis platform designed for enterprise-scale operations and complex data workflows.
Visit Datagrid →

Model Updates

OpenAI o1 Reasoning Model Updates
OpenAI has released updates to its o1 reasoning model, improving performance on complex mathematical and scientific problems. The enhanced model shows significant improvements in multi-step reasoning tasks.

Claude 3.5 Sonnet Computer Use Beta
Anthropic has released a beta version of Claude 3.5 Sonnet with computer use capabilities, allowing the AI to interact directly with computer interfaces and perform complex, multi-step tasks.

Alibaba's Qwen3 Coder Release
Alibaba has released Qwen3 Coder, described as the most agentic code model in the Qwen series. The model is available in multiple sizes including a 480B parameter version.
Explore Qwen3 →

Mistral Codestral 25.08 Enterprise Platform
Mistral AI has unveiled Codestral 25.08 as part of a complete enterprise coding stack. The update includes 30% increase in accepted completions and 50% fewer runaway generations.
Try Codestral →

Gemini 2.5 Pro Enhanced Access
Google has significantly increased limits for Gemini 2.5 Pro access, particularly for education users who now receive free access to the advanced model.
Access Gemini 2.5 Pro →

Google Launches Gemini Deep Think AI
Google has launched Gemini Deep Think, a multi-agent reasoning model designed for complex problem-solving scenarios and advanced analytical tasks.
Learn more →

New Anthropic Research: Persona Vectors
Anthropic has published new research on persona vectors, exploring how AI models can maintain consistent personality traits and communication styles across different interactions.
Read the research →

Other things we’re loving

GenAI Apps Double Revenue to 1.7 Billion Downloads — Generative AI applications have doubled their revenue and reached 1.7 billion downloads in the first half of 2025, with users spending 15.6 billion hours on these apps, demonstrating massive consumer adoption.
TechCrunch

YouTube Launches AI Age Detection Technology — YouTube has introduced AI age detection technology in the US to identify minors regardless of entered birthdate, automatically applying restricted advertising and wellbeing tools for under-18 users.
The Verge

Apple Loses Fourth AI Researcher to Meta — Apple has lost its fourth key AI researcher, Bowen Zhang, to Meta's superintelligence team within a month, highlighting the intense competition for AI expertise and Meta's aggressive recruitment strategy.
Bloomberg

Amazon Invests in Fable's "Netflix of AI" — Amazon has invested in Fable, a startup developing Showrunner, an AI-generated TV show service that allows users to create scenes or entire episodes using prompts, representing innovation in AI-generated entertainment.
Variety

Anthropic Raises £3-5 Billion at £170 Billion Valuation— Anthropic is reportedly raising £3-5 billion at a £170 billion valuation, nearly tripling its March valuation and reflecting massive investor confidence in the company's Claude AI models.
TechCrunch

Stanford Creates Virtual AI Laboratory — Stanford University has developed a virtual AI laboratory that accelerates scientific discovery by simulating complex experiments and research processes, potentially revolutionising the pace of scientific research.

AI Agents Struggling with Real-World Tasks — Recent studies reveal that AI agents are failing at basic real-world tasks, with success rates significantly lower than expected despite advances in language models.

OpenAI Reaches £12 Billion Annualised Revenue — OpenAI has doubled its revenue in the first seven months of 2025, reaching £12 billion in annualised revenue with approximately 700 million weekly active ChatGPT users.

Your Public ChatGPT Queries Are Getting Indexed by Google — Public ChatGPT conversations are now being indexed by search engines, raising privacy concerns about AI-generated content appearing in search results.
TechCrunch

Tim Cook Defends Apple's AI Delay — Apple CEO Tim Cook defended the company's measured approach to AI development, stating "We've rarely been first" but emphasising quality over speed in AI implementation.
MacRumors

Dramatic Rise in Construction AI Adoption — The construction industry is experiencing dramatic AI adoption growth, though the sector must bridge a significant "maturity gap" between early adopters and mainstream implementation.
PBC Today

New York Startup Raises £40 Million for Construction AI — A New York startup has raised £40 million in funding to develop AI agents for handling administrative tasks in the construction industry, demonstrating growing investor interest in sector-specific AI solutions.
Business Insider

Anthropic Researchers Discover AI Thinking Paradox— Anthropic researchers have discovered that longer thinking time can sometimes make AI models perform worse, revealing unexpected complexities in AI reasoning processes.
Ground News

Google's NotebookLM Rolls Out Video Overviews — Google has expanded NotebookLM with video overview capabilities, allowing users to generate video summaries from their research materials and documents.
TechCrunch

Community

The Spotlight Podcast

How Pascal + Watson Navigates the AI Revolution in Architectural Practice

What happens when a leading architectural firm takes a methodical approach to AI adoption? In this thought-provoking conversation with Pavan Birdi, Senior Associate at Pascal + Watson, we explore the strategic implementation of artificial intelligence in architectural practice—particularly within risk-sensitive sectors like aviation and defence.

Listen Now:
Apple

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