This Week’s BIG thing

Summary

OpenAI and Anthropic have forged significant agreements with the US government to advance AI safety and transparency. These partnerships involve granting the US AI Safety Institute access to their most sophisticated AI models for thorough testing and evaluation, aiming to mitigate risks associated with AI misuse and unforeseen consequences. As reported by Reuters, this collaboration is poised to set new standards in responsible AI development. Furthermore, Anthropic has demonstrated a commitment to openness by releasing detailed system prompt guides, enhancing transparency and understanding of their AI systems' inner workings.

Why You Should Care

For professionals in project delivery sectors such as construction, energy, and infrastructure, these developments are particularly pertinent. AI technologies are increasingly integral to optimising project management, enhancing efficiency, and reducing costs. The collaboration between OpenAI, Anthropic, and the US government ensures that these AI tools are not only powerful but also safe and reliable. According to The Verge, this initiative will help "prevent potential abuses and ensure AI systems operate within ethical and secure boundaries." Such assurances are essential for building trust and facilitating broader adoption of AI solutions in complex projects, ultimately driving productivity and innovation across the industry.

 The rabbit hole 🐰

To explore this topic further, consider examining how these agreements might influence global AI regulatory frameworks and standards. The detailed disclosures in Anthropic's system prompts documentation offer a unique opportunity to understand the nuances of AI operations, which could inform best practices in your projects. Additionally, analysing the potential impacts of enhanced AI safety measures on future project delivery methodologies can provide valuable insights. Will increased oversight and transparency lead to more robust and adaptable AI applications tailored to the specific needs of industries like real estate and property development? Delving into these questions can help you stay ahead in leveraging AI for effective and secure project outcomes.

What’s new: Projects

Surprising ChatGPT Stat: Adoption Amongst Researchers is Low

A recent study by CBRE reveals that the adoption of generative AI tools, such as ChatGPT, among researchers is surprisingly low. The study indicates that fewer than 10% of researchers use generative AI tools regularly, while over half have never used them at all. This finding is striking, given the rapid advancement and availability of AI technologies that can significantly enhance research processes. The report delves into potential reasons for this low adoption rate, including concerns over data privacy, the accuracy of AI-generated outputs, and a general lack of familiarity with the tools.

Why should you care? 🫵

This finding has significant implications for industries that rely heavily on research and data analysis, such as real estate, construction, and infrastructure. AI tools like ChatGPT have the potential to revolutionise how research is conducted by automating data analysis, generating insights, and even predicting trends. However, the CBRE study highlights a major gap in AI utilisation that could be limiting the efficiency and innovation of research outputs. For professionals in project delivery, understanding and integrating generative AI tools could provide a competitive edge, enhancing the quality and speed of research-related tasks, from market analysis to project feasibility studies.

 The rabbit hole 🐰

To dive deeper into this issue, consider exploring the reasons behind the slow adoption of generative AI tools in research. The CBRE article suggests that concerns over data privacy and the reliability of AI-generated content are key barriers. Investigating these concerns could provide insights into how to address them, potentially leading to broader adoption of AI tools in your own projects. Additionally, examining successful case studies where generative AI has been effectively integrated into research workflows might offer practical strategies for overcoming resistance and maximising the benefits of AI in your field. Could this technology, if better understood and trusted, unlock new efficiencies and innovations in project delivery, from initial research phases to final execution?

💙 Other project news we’re loving

• ^{Could AI Visualisation Solve Britain's Great Retrofit Challenge?
  AI visualisation tools could be key to overcoming the UK's retrofit challenge, enhancing decision-making and efficiency in sustainable building projects.
  Read more}

• ^{UK's First Teacherless AI Classroom Set to Open in London
  A pioneering AI-powered classroom without teachers is opening in London, revolutionising education by using AI to personalise learning for students.
  Read more}
  

What’s new: Productivity

 Where is ChatGPT’s version of Google?

“We think there is room to make search much better than it is today”, this is what Sam Altman had to say about the current state of internet search in reference to OpenAI’s hotly anticipated SearchGPT. So what is SearchGPT? When is it out and how has early testing being received?

Search GPT’s Features

SearchGPT will be the evolution of ChatGPT, where your questions will be met with answers coupled with clear source references and summarisations. Features include:

• A ChatGPT like interface that lets users ask follow-up questions while keeping the context intact. Simple, familiar, nice.

• Summarised answers drawn from multiple sources on the web, with links to the original content. This feature is key to sense-checking the outputs from LLM’s which often hallucinate information!

• Integration of images and possibly other media related to the search. The visual element was a noticeable absence in ChatGPT, how they navigate licensing is another legal hurdle big tech keep stumbling over.

• Collaborations with leading publishers to showcase their content. We’re interested to see how balanced content will be, as collaborations can lead to biased content pushed by SearchGPT. Public perception hasn’t taken to this too well.

What do early testers think of it?

SearchGPT was first introduced as a prototype in a limited release to 10,000 test users on July 26, 2024. Early reviews from select testers have been mostly positive, with users valuing SearchGPT’s ability to compile information into summaries and lists. However, some concerns remain about citation accuracy, especially on complex topics. A YouTube review highlighted its strength in quickly summarising content, though some testers feel it lacks the depth Google offers for in-depth research, and has a greater tendency to hallucinate than Perplexity.

Potential Impact on Internet Search

While unlikely to replace Google in the near term given Google's dominant market share, experts suggest SearchGPT could potentially capture a notable 5-10% of the search market by offering an efficient, ad-free experience with direct answers. Even a modest shift would significantly disrupt the search industry.

No release has been given but the waitlist has been closed.

Why should you care? 🫵

The way in which we consume information is going to significantly change over the next few years. AI advancements like SearchGPT, bring personalisation to a new level, supercharging our ability to learn. Search providers excel in different things, our advice to you is to firstly understand what you want from the search and then go with the provider that does that job better. E.g. SearchGPT, much like GPT will excel at simplifying concepts into easy to digest chunks, whereas it might have a tendency to hallucinate and not provide reliable source references like Perplexity. Equally, Google still provides the most balanced and flexible search allowing you enter deeper rabbit holes with the comfort of a familiar interface.

The rabbit hole 🐰

🔗 Hear from OpenAI on ChatGPT’s SearchGPT

🔗 A great review and comparison of SearchGPT to other competitors

🔗 A data-driven comparison of search engines

🔗 Concerns of delays, becoming typical of OpenAI

🔗 A dip into the market share of search

💙 Other Project news we’re loving

•  ^{Google AI Overviews Now Available in the UK
  Google has expanded its AI-powered search overviews to the UK, enhancing how users find information with AI-generated summaries and insights.
  Read more} 

• ^{Claude Anthropic Artifacts Now Available Everywhere
  Anthropic has made its AI-powered Claude Artifacts globally accessible, providing advanced tools for AI research and development.
  Read more}
  

What’s new: Tech

Amazon’s new 'Remarkable' Alexa: A game-changer in voice assistants with Claude AI integration

This year, Amazon are set to roll out a significantly upgraded version of its Alexa voice assistant, now powered by the frontier LLM model Claude AI from Anthropic. This marks a major pivot for Amazon, after recognising the limitations of its in-house AI, they are competing on all fronts, by partnering with the AI giant who are also leading the way in AI transparency.

What’s New and Exciting?

• Conversational depth and memory: Alexa will engage in more fluid and context-aware conversations, remembering details from past interactions to enhance user experience. Memory is king!

• Complex task handling: From ordering dinner to drafting emails, Alexa will be able to manage intricate tasks through a single voice command. We are skeptical of this, as AI hasn’t done that well in being able to execute end-to-end tasks like ordering food historically.

• Tailored recommendations: Expect highly personalised suggestions for shopping, news, and entertainment based on your habits and preferences.

• Smarter home automation: Alexa will better understand and anticipate your needs, remembering preferences to automate tasks like adjusting the lights or playing your favourite music. This connection to smart homes, brings a new level of interaction with AI.

• Detailed Assistance: Whether you're shopping for a specific item or need advice on a complex query, Alexa will provide in-depth guidance and recommendations.

A Much Needed Pricing Model: Free vs. Premium

Amazon plans to introduce a subscription model for the ‘Remarkable’ Alexa, with monthly fees expected to range from $5 to $10. This premium service will run alongside a free ‘Classic’ version of Alexa, which will continue to offer basic features.

This move to a paid model reflects Amazon's strategy to recoup some of the significant investments made in developing Alexa. However, with consumers accustomed to free AI services, it remains to be seen whether this approach will succeed.

Competing in the AI-Assistant Arena

With this partnership, Amazon positions itself to compete with other tech giants who are also integrating advanced AI into their smart assistants:

• ^{Apple is embedding OpenAI’s ChatGPT into Siri as part of its new Apple Intelligence suite in iOS 18 and macOS Sequoia. (Supposed release Q4 2024)}

• ^{Google has launched its AI assistant, Gemini, which comes with a $20 monthly subscription for advanced features. (Already launched)}

Why should you care? 🫵

LLM’s being integrated into AI assistants like Alexa and Siri are set to transform how we interact with technology in our daily lives and at work. For everyday uses, this means more natural and conversational interactions with AI, making it easier to complete complex tasks like scheduling or shopping with just a simple voice command. At the same time, for project professionals, getting project insights via voice offer’s a new dimension for accessing project information. We’re often blinded by heavy visual reports, that this approach is refreshing and in cases can improve communication much like getting clarification via a call after reading an ambiguous email. As these AI assistants become more capable, they’ll not only simplify routine tasks but also support more informed, efficient, and personalised workflows, blending seamlessly into both our personal and professional lives.

The rabbit hole 🐰

🔗 An intro by Reuters, Why Amazon turned to Anthropic’s Claude

🔗 Ever wondered why its taken so long for voice assistants to incorporate LLM’s?

🔗 Why this development is the key for the survival of Alexa

🔗 Apples similar movements to come later this year

https://www.macrumors.com/2024/08/30/amazon-launching-new-alexa-claude-ai/

💙 Other tech news we’re loving

• ^{🌎 Will AI be humanities final invention? An excellent high-production video}

• ^{⛏️ AI simulated civilisation in Minecraft}

• 🌠 ^{Midjourney teasing hardware}

• 👾 ^{Google AI creating personalised generative gaming: DOOM!}

• 🤖 ^{What do LLM’s picture themselves to look like?}

• ^{Perplexity AI to Start Running Search Ads in Q4
  Perplexity AI plans to introduce search ads by the fourth quarter of 2024, marking its entry into the competitive AI-powered search market.
  Read more}

• ^{Eric Schmidt’s AI Prophecy
  Former Google CEO Eric Schmidt shares his vision for AI’s future, predicting profound societal impacts and the rise of AI-driven decision-making.
  Read more}

• ^{Cerebras Unveils Instant-Speed AI Inference
  Cerebras introduces its latest AI technology, promising instant-speed inference capabilities that could revolutionise AI applications across industries.
  Read more}

• ^{OpenAI Valuation Set to Surpass $100 Billion with Potential Investments from Apple and Nvidia
  OpenAI is in talks to raise funding that could value it at over $100 billion, with tech giants Apple and Nvidia reportedly considering investments to capitalise on AI's growing influence.
  Read more}

Events

Oprah to Interview Sam Altman and Bill Gates

On 12th September, Oprah Winfrey will conduct a highly anticipated interview with Sam Altman, CEO of OpenAI, and Bill Gates, the co-founder of Microsoft. This special event will delve into the future of AI, its potential impact on society, and the ethical considerations surrounding its rapid development. Altman, who leads one of the most influential AI companies, and Gates, a long-time advocate for technology-driven solutions, will share their insights on how AI could shape the future of industries, including healthcare, education, and beyond.

Event Details:

• Date: 12th September

• Time: The exact time of the broadcast has not been confirmed yet, so keep an eye out for updates on major platforms.

• Where to Watch: The interview will be available across various streaming platforms and is expected to be widely covered by media outlets.

This conversation promises to be a thought-provoking discussion on AI's evolving role in our world, making it a must-watch for professionals across all industries.

Project Flux Podcast

This week’s episode features Antony Slumbers, a renowned expert in proptech and space-as-a-service. The discussion delves into the integration of generative AI in the real estate sector, drawing parallels between today’s AI advancements and past technological shifts. Slumbers also explores how leaders can balance current demands with future needs, the necessity for new business models, and the startup opportunities in the AI era. The conversation touches on AI’s potential to augment rather than replace human creativity, especially in real estate and construction.

Course Information: Generative AI for Real Estate - Next cohort starts 6th September 2024.

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…