AI-Powered Research: OpenAI’s Deep Research Changes the Game

Table of Contents

• OpenAI’s Deep Research: AI-Powered Reports in Minutes

• AI Wars: US vs China – The Battle Heats Up

• Presented by Cogram

• Why Big Projects Keep Failing—And How AI Might Help (or Not?)

• U.S. Copyright Office Clarifies Rules for AI-Generated Content

• The AI Paradox in Projects by Yoshi Soornack

• A Tale of Two Futures: AI’s Transformation of Project Management by James Garner

• Podcast of the week

• The AI Cycle: From Innovation to Enslavement—Is Humanity Doomed to Repeat History?

• Tip of the Week

• Meet Project Flux: About Us

OpenAI’s Deep Research: AI-Powered Reports in Minutes

OpenAI has unveiled Deep Research, an advanced AI agent capable of researching, analysing, and synthesising data from hundreds of online sources in just 10 minutes. Unlike standard AI summarisation tools, Deep Research acts as a dynamic investigator—adjusting its approach, bypassing paywalls, and uncovering novel connections. This innovation promises to revolutionise research-heavy industries. Read more here.

A recent performance chart on GPQA Diamond showcases OpenAI’s exponential growth in AI research capabilities, with Deep Research and GPT-4o Pro surpassing human PhDs in certain tasks. This trend suggests that AI-powered research is not just evolving—it’s accelerating at an unprecedented rate.

What Does This Mean for Me?
For professionals in project delivery, Deep Research could be a game-changer. Whether you’re assessing risks in construction, analysing market trends in real estate, or tracking sustainability regulations in energy, AI-powered research could cut hours of work into minutes. This means faster decision-making, more accurate reports, and a competitive edge in fast-moving industries.

Key Themes or Details

• AI-Powered Research – Deep Research automates the work of skilled researchers, compiling high-quality reports in minutes.

• Adaptive Investigation – The tool refines its research approach dynamically, overcoming obstacles like paywalls and information gaps.

• Exponential Growth – Performance data suggests AI research capabilities are rapidly accelerating, outpacing traditional research methods.

• Industry Impact – Project managers, analysts, and consultants can leverage AI-driven insights for better project planning and risk management.

• Time Efficiency – Reports that once took hours now take minutes, freeing up professionals for strategic tasks.

Down the Rabbit Hole

• OpenAI’s official announcement: Read here

• Deep Research in action – Lifehacker’s take: Check it out

• OpenAI’s Twitter announcement: See the post

• Sam Altman’s thoughts: Tweet 1 | Tweet 2

• Performance Chart on GPQA Diamond: See below

What’s new: Tech

AI Wars: US vs China – The Battle Heats Up

Big Tech Moves and Policy Shifts

The AI arms race is accelerating. Amazon is set to unveil its AI-powered Alexa on 26 February, potentially introducing a subscription model. Google is expanding its Gemini 2.0 suite with Flash-Lite for efficiency and Pro Experimental for advanced reasoning. In a controversial shift, Google has also removed its AI weapons restriction from its ethical principles, raising concerns over future military applications. Meanwhile, a new US bill could criminalise downloading Chinese AI models like DeepSeek, with penalties reaching $1M in fines or 20 years in jail.

What Does This Mean for Me?

This AI battle has direct implications for businesses and professionals:

• Project Delivery & Productivity – AI assistants like Alexa and Gemini could streamline workflows, but a subscription model might mean higher costs for businesses.

• Regulatory Uncertainty – If the US bill passes, companies must tread carefully when using AI tools from China, potentially disrupting global collaboration.

• Ethical Dilemmas – Google’s shift on AI ethics could spark debates on AI’s role in defence and security projects.

Key Takeaways:

• Amazon’s AI Alexa launch on 26 Feb could introduce a subscription model.

• Google expands Gemini 2.0, adding Flash-Lite and Pro Experimental.

• Google removes AI weapons ban from its ethical principles.

• US bill may criminalise Chinese AI use, risking major fines or jail time.

Down the Rabbit Hole

• Google AI Principles Update

• Amazon’s Alexa AI Rumours (Check closer to 26 Feb)

• US Bill Targeting Chinese AI

💙 Other tech news we’re loving

• 🚀 _{Apple’s Robot Revolution? - Apple is reportedly prototyping a tabletop AI-powered robotic assistant, hinting at a future where AI-driven devices handle everyday tasks. Could this reshape smart home automation? Read more.}

• 🧠 _{OpenAI’s O3-Mini Gets Transparent - OpenAI’s O3-mini model now reveals its “thought” process before answering, improving user trust and interaction. This shift aims to make AI decisions more understandable. Learn more.}

• _{⚠️ Is AI Labelling Misleading? - A new study questions how AI is marketed, revealing public confusion over AI’s actual capabilities. Transparency is key for businesses using AI branding. Explore the study.}

• 💡 _{OpenAI’s Future Plans Leaked - A US trademark filing hints at potential new OpenAI products—whether real or speculative remains to be seen. Meanwhile, OpenAI and SoftBank launch SB OpenAI Japan. Details here.}

• 🔎 _{Perplexity Integrates DeepSeek R1 - Perplexity now supports DeepSeek R1, offering uncensored AI responses on US/EU servers. Pro users get 500 searches daily, while free users receive five. Try it now.}

• 🔄 _{OpenAI’s Leadership Shuffle - Co-founder John Schulman is rumoured to be joining ex-CTO Mira Murati’s next startup, shaking up AI leadership. Find out more.}

• 🤖 _{Figure AI Breaks Away from OpenAI - Robot startup Figure AI is forging its own path, claiming a breakthrough independent of OpenAI. A major move in the AI robotics space. More here.}

Presented by Cogram

Cogram is an AI assistant specifically designed for AEC to help professionals save time on documentation and prevent disputes. 

• Automated meeting minutes: Cogram drafts meeting minutes directly in your company’s template with tailored levels of detail, format, and style.

• Works for virtual and in-person meetings: Cogram works across Zoom, Teams, and in-person meetings, such as site inspections and kick-off meetings. 

• Field reports in minutes: Cogram’s mobile app combines site photos with minutes from on-site discussions to draft field reports in minutes.

• Project-wide insights: Ask questions about your projects and get answers drawn from every relevant meeting and email chain from that project for the most up-to-date information. 

Learn more and schedule a 15-live demo of Cogram.

What’s new: Projects

Why Big Projects Keep Failing—And How AI Might Help (or Not?)

Professor Bent Flyvbjerg, the world’s leading expert on megaproject failures, has unleashed a LinkedIn post on AI’s role in project management. He criticises the hype around AI tools like ChatGPT and Perplexity, arguing that data quality—not AI—is the real issue. AI can’t fix projects if the underlying data is flawed, biased, or incomplete.

Instead of chasing AI as a silver bullet, Flyvbjerg insists we should focus on better project data, historical benchmarks, and rigorous planning. Without that, AI will just make bad decisions faster.

What Does This Mean for Me?

• AI alone won’t save failing projects – Clean, accurate data is king.

• Bad project planning + AI = faster disasters – AI can optimise but not correct fundamental project flaws.

• Historical benchmarks matter – The best decisions come from learning from past project failures and successes.

Key Takeaways:

• Bent Flyvbjerg argues AI is overhyped in project management.

• Flawed data = flawed AI decisions – focus on improving data quality.

• AI should assist, not replace, solid project planning.

• Megaproject failures are predictable—use historical data and benchmarks.

Down the Rabbit Hole

• Bent Flyvbjerg’s LinkedIn (Post 1)

• Bent Flyvbjerg’s LinkedIn (Post 2)

What’s your take? Is AI helping or harming project delivery? 🚀

💙 Other Project news we’re loving

• _{🏗️ AI Meets BIM: The Future of AEC? - OpenAI’s integration with BIM could transform architecture, engineering, and construction, enabling smarter decision-making and streamlined workflows. Is AI finally ready to revolutionise project delivery? Read more.}

• _{💰 Microsoft’s Copilot Price Standoff - Despite buyer reluctance, Microsoft refuses to lower Office 365 Copilot’s per-seat price, as businesses weigh AI’s value against budget constraints. Is Copilot worth the cost? Find out more.}

• _{⚖️ Anthropic’s AI for Legal and Tax Pros - Anthropic and Thomson Reuters built a trustworthy AI for tax and legal professionals, tackling accuracy concerns in high-stakes industries. Discover the details.}

• 🔎 _{Google’s Gemini 2.0 Pro Experimental Leads AI Race - Google’s Gemini 2.0 Pro Experimental tops LMArena rankings, while its Flash Thinking model challenges OpenAI and DeepSeek in advanced reasoning. Explore Gemini’s power.}

• 🚨 _{DeepSeek R1 Flunks AI Safety Test - DeepSeek’s open-source model failed all 50 safety prompts in the HarmBench dataset, raising concerns about AI’s ability to handle harmful queries. Read the report.}

• _{🛍️ AI-Powered Virtual Stylist Boosts Sales by 30% - A new AI home decor stylist personalises shopping experiences, helping customers find the perfect items—driving a 30% sales increase. See how AI is transforming retail.}

• 🩺 _{UK’s Major AI Breast Cancer Screening Initiative - The NHS is launching AI-powered breast cancer screenings, aiming for faster, more accurate diagnoses. Could AI revolutionise early detection? Read more.}

• _{🛣️ AI and Self-Healing Roads? - AI-driven self-healing road technology could detect cracks before they worsen, saving billions in repairs. The future of smart infrastructure is here. Explore the innovation.}

What’s new: Productivity

U.S. Copyright Office Clarifies Rules for AI-Generated Content

The U.S. Copyright Office has released Part 2 of its landmark report, Copyright and Artificial Intelligence, focusing on whether works created using generative AI qualify for copyright protection. Here’s what you need to know:

Human Authorship Remains Essential

The report reaffirms that purely AI-generated content—created without meaningful human control—cannot be copyrighted. For example, a text prompt like “cartoon spaceship” or even a detailed description fed into an AI tool does not, by itself, grant authorship rights to the user.

When Can AI-Assisted Works Be Copyrighted?

Protection is possible if people contribute creative expression:

• Modifications/Arrangements: Editing AI-generated outputs (e.g., adjusting images, compiling text) may qualify if choices meet the originality standard.

• Expressive Inputs: If a user provides copyrighted material (e.g., a hand-drawn sketch) that is perceptible in the final AI output, that portion is protected.

• Compilations: Curating AI-generated content into a creative arrangement (e.g., a comic book blending AI art with original text) can be copyrighted as a compilation.

Prompts Alone Are Not Enough

The report emphasises that prompts are ideas, not authorship. Even detailed prompts (e.g., specifying lighting, style, or themes) do not guarantee control over the AI’s output. Effectively, the manner of execution determines authorship. For instance, China’s courts recently protected an AI-generated image because a user made 150+ iterative prompts and edits, demonstrating “intellectual achievement.”

Key Takeaway

AI is a tool, not an author. To secure copyright, focus on meaningful human creativity—whether through editing, compiling, or integrating original work.

Read the full report: Copyright and Artificial Intelligence, Part 2

Want more insights on AI and IP? Subscribe for updates at www.copyright.gov/ai.

💙 Other productivity news we’re loving

• _{🛠️ Arcadis Empowers Staff with AI Innovation - Arcadis and Microsoft have developed a platform enabling employees to build Power Platform solutions, leveraging AI-driven chatbots to boost project-level digital innovation. Explore the platform.}

• _{💰 Musk Turns to AI for Cost-Cutting - Elon Musk is deploying AI to identify budget savings in government and business, aiming for efficiency through automation. But will AI-driven cuts go too far? Read more.}

• 🇪🇺 _{OpenAI Introduces EU Data Residency - To comply with strict EU regulations, OpenAI now offers European data residency, giving businesses greater control over sensitive information. Learn more.}

• _{🌍 Europe Unites to Challenge US AI Dominance -European tech giants are collaborating to rival US and Chinese AI dominance, focusing on sovereign AI solutions. Can Europe compete? Find out more.}

• 🚫 _{EU Bans ‘Unacceptable’ AI Systems - The EU has outlawed AI systems posing an unacceptable risk, reinforcing its stance on AI safety and ethics. Read the policy update.}

• 📈 _{AI Strategies for Viral Content - LinkedIn influencer Ruben Hassid reveals AI-driven content strategies, emphasising taste, experience, and high standards to stay ahead in digital marketing. Watch here.}

• 🤖 _{The Rise of Humanoid Robots - Elon Musk calls humanoid robots “the biggest product in history,” predicting they’ll soon outnumber humans. Nvidia’s CEO agrees: “Everything that moves will be robotic.” Read more.}
  

The AI Paradox in Projects by Yoshi Soornack

As we step into 2025, the rapid advancement of AI continues to reshape industries, offering unprecedented opportunities—and challenges. While AI has the potential to revolutionise projects across sectors, its growing complexity and ethical dilemmas have left many hesitant to fully embrace it.  

The Complexity of AI

AI models like GPT-4o, ChatGPT’s o1, and DeepSeek-R1 are becoming more powerful but also more opaque and unpredictable. Key challenges include:  

- Robustness: Sensitivity to edge cases can lead to destabilising decisions.  

- Interpretability: Lack of transparency undermines trust and accountability.  

- Control: Autonomous systems may prioritise speed over safety.  

- Ethicality: AI risks amplifying human biases and promoting harmful outcomes, like burnout.  

The AI Paradox  

Large projects carry significant risks, especially when combined with AI’s unpredictability—evidenced by failures like Zillow’s $304m write-down. Smaller projects, while less risky, limit AI’s impact, creating a dilemma for organisations. This creates an AI paradox.

Aligning AI with Human Values

To build AI effectively, efforts today aim to align it with human intent, which is critical but challenging. AI mirrors our flaws, codifying biases from historical data. Without careful alignment, AI risks perpetuating systemic discrimination and unethical practices.  

A Triadic Approach to Effective AI Integration

☕️ Get actionable steps in our full story, so that you can implement AI safely on your projects.

A Tale of Two Futures: AI’s Transformation of Project Management by James Garner

AI is reshaping project delivery, offering both unprecedented efficiency and disruptive change. Like Dickens’ A Tale of Two Cities, we stand at a crossroads—between enlightenment and uncertainty. How can project managers navigate this transformation? Read more to explore the alchemy of AI and human insight.

Podcasts of the Week

🎙️ Project Flux Podcast: AI, Innovation & Project Delivery

In this insight-packed episode, Dale Fung, founder of Project Chatter, discusses AI’s role in project controls, innovation in leadership, and the rise of agentic AI. He shares how professionals can navigate the evolving landscape and embrace AI as a thought partner. A must-listen for project managers! Listen here. 🚀

🎓 AI Explained in 3 Hours

Want to truly understand AI? Andrej Karpathy, former OpenAI and Tesla AI lead, delivers a masterclass on how ChatGPT and AI models work. From token prediction to AI’s real-world applications, he makes complex concepts accessible. A must-watch for professionals embracing AI! Watch here.

The AI Cycle: From Innovation to Enslavement—Is Humanity Doomed to Repeat History?

Tip of the Week

💡 Tip of the Week: Ask AI the Right Question

AI isn’t just for answers—it’s a powerful tool for self-reflection. Try this experiment: Ask AI, “Knowing what you know about me, what’s the most important question I should ask myself?”

This prompt encourages deeper thinking, helping you uncover insights about your goals, values, and direction. Use it to refine decision-making, challenge assumptions, and spark creativity in both work and life. Stay curious, and let AI be your thought partner! 🚀

Try it out and share your biggest takeaways! #AI #SelfReflection #ProjectFlux

Meet Project Flux: About Us

At Project Flux, we're committed to pioneering the future of construction and project delivery through the lens of cutting-edge Artificial Intelligence insights. Our vision is to be at the forefront of integrating AI into the fabric of project delivery, transforming how projects are conceptualised, planned, and executed.

Our Mission Project Flux aims to not only inform and educate but also to inspire professionals in the construction industry to embrace the transformative potential of AI. We believe in the power of AI to revolutionise project delivery, making it more efficient, predictive, and adaptable to the dynamic demands of the modern world.

What We Offer Through our insightful newsletters, podcasts and curated content on LinkedIn, and engaging discussions, Project Flux serves as a resource for professionals seeking to stay ahead in their field. We offer a blend of practical advice, thought leadership, and the latest developments in AI and construction technology.

What People Say…