The ChatGPT Observer #25

If you're short on time, here's your one-minute summary of this week's key AI news and advancements:

?? Top Pick: OpenAI's ChatGPT Enterprise enhances data analysis capabilities, offering features like direct data warehouse querying and code generation.

?? This Week's Spotlight: EY's AI Adoption Survey reveals significant trends in AI investment and adoption across industries.

?? Key Statistics: 74% of IT professionals fear AI could make their skills obsolete, highlighting the need for upskilling.

?? ChatGPT & OpenAI News: OpenAI partners with Condé Nast for content creation, launches GPT-4o fine-tuning, and Upwork integrates OpenAI models.

? Claude Sonnet News: Anthropic adds LaTeX rendering to Claude and surpasses $1M in mobile app revenue.

?? Humanoid Robots: Agibot unveils robots to rival Tesla's Optimus, Boston Dynamics shows Atlas robot doing push-ups, and iRonCub3 introduces jet-powered humanoid.

?? Corporate AI Developments: ElevenLabs launches global text-to-speech app, Microsoft unveils PHI-3.5 model, and Midjourney updates web platform.

?? AI Innovations: New tools emerge for real-time face swapping, text-to-video generation, and AI-powered 3D modelling.

?? AI in Business: Best Buy introduces AI-powered delivery tracking, Google develops cough disease detection model, and Amazon Cloud CEO predicts AI will transform software engineering.

? Responsible AI: McAfee integrates deepfake detector in Lenovo PCs, authors join copyright lawsuit against OpenAI, and Hong Kong issues generative AI guidelines for finance.

?? For more details, read the full newsletter!

?? Top Pick: ChatGPT Enterprise for Data Analysis ??

OpenAI's webinar "Enabling a Data-Driven Workforce with ChatGPT Enterprise" offers valuable insights into how ChatGPT Enterprise can transform organizational data analysis capabilities.

Key highlights include:

1. Data Ownership and Control: Your business data remains yours. OpenAI doesn't train on it, and you have full control over data retention.
2. Customisable Access: Features like SAML SSO and role-based access control ensure you decide who accesses what.
3. Robust Security: With CCPA, CSA STAR, and SOC 2 Type 2 compliance, plus support for HIPAA, FINRA, and GDPR, your data is in safe hands.
4. Cross-Departmental Applications: From marketing campaign analysis to financial debt analysis and operational forecasting, ChatGPT Enterprise has wide-ranging use cases.
5. Direct Data Warehouse Querying: Users can query complex datasets using natural language, democratising data analysis across the organisation.
6. Code Generation: ChatGPT Enterprise can generate code in major programming languages and write SQL queries.

Practical Example: Users can ask questions like: "List all flights from NYC's major airports in January 2013. Include the flights information + origin + destination airport information" and receive a comprehensive CSV file and report with visualisations.

The ChatGPT Enterprise effectively combines the roles of a data analyst and visualization expert, making advanced data analysis accessible to a broader range of employees.

For a detailed demonstration of these features, you can watch the full webinar (42 minutes) here:

?? This Week's Spotlight: EY's AI Adoption Survey

EY's recent survey of 500 senior executives across various industries reveals the following trends in AI adoption:

?? AI Investment

• AI investments are projected to nearly double, with many companies planning to invest over $10 million in AI next year.
• 88% of senior leaders currently allocate 5% or more of their total budget to AI.
• 50% of senior leaders expect to dedicate 25% or more of their total budget to AI investments in the coming year.

?? Return on AI Investments

• 77% of organisations investing in AI reported positive returns in operational efficiencies.
• 74% of organisations saw increased employee productivity due to AI investments.
• 72% of organisations reported improved customer satisfaction as a result of AI initiatives.

companies that do not actively engage with AI risk being left behind in a market that increasingly rewards innovation and agility

??? AI Maturity and Infrastructure

• 34% of senior executives say their organisation is aligning AI strategy with business objectives fully and at scale.
• 36% of senior leaders report that their organisation is investing in data infrastructure fully and at scale.
• 35% of senior leaders say their organisation is creating a roadmap for AI implementation fully and at scale.

?? Responsible AI Adoption

• 53% of senior leaders have reported increased organisational interest in responsible AI over the past year.
• 34% of organisations are building an AI governance framework fully and at scale.
• 32% of organisations are addressing bias in AI models fully and at scale.

?? Talent Development

• 40% of senior leaders are encouraging employees to embrace AI fully and at scale.
• 37% of senior leaders report that their organisation is training or upskilling employees on AI fully and at scale.
• 78% of organisations struggle to find employees with the necessary AI skill set.

This advancement goes beyond simple task improvement — it calls for a radical redesign of business processes to be AI-centric. By doing so, companies are not just automating; they are innovating, so that their investments in AI yield measurable financial returns and solidify their standing for the future.

Key Strategies for AI Adoption:

1. Adopt diversified AI investment strategies: Balance custom AI development with the acquisition of ready-made AI products.
2. Prioritize ROI-driven AI deployment: Focus on AI tools that enhance operational efficiency and employee productivity.
3. Align your business and AI maturity roadmap: Integrate AI initiatives with core business objectives and build a robust data infrastructure.
4. Invest in responsible AI as a competitive edge: Adopt ethical AI standards to foster trust, preempt regulatory challenges, and gain a competitive advantage.
5. Embrace talent development as a value driver: Invest in upskilling and reskilling programs to develop an AI-fluent workforce.

Read more: EY AI Adoption Survey

?? Key Statistics: AI Adoption Sparks Concern Among IT Professionals About Skill Obsolescence

A recent survey highlights growing concerns among IT professionals about the impact of AI on their careers.

• 74% of IT professionals believe AI advancements could make their skills obsolete.
• 87% of respondents are concerned about the rapid pace of AI development in the tech industry.
• 64% are actively seeking opportunities to upskill and adapt to AI technologies.
• 56% of IT pros feel that their organisations are not providing adequate training to cope with AI-driven changes.

Read more: CIO

?? ChatGPT & OpenAI News

• OpenAI and Condé Nast Partner to Revolutionise Content Creation ??: OpenAI has teamed up with Condé Nast to integrate AI into the media giant's content operations. This partnership aims to enhance content production, improve audience engagement, and streamline workflows using advanced AI technology, showcasing the growing role of AI in the media industry. Read more: OpenAI
• Fine-Tuning Now Available for GPT-4o ???: OpenAI has launched fine-tuning for GPT-4o, enabling developers to achieve higher performance and accuracy by customising models with their datasets. Companies like Cosine and Distyl have already seen impressive results, with Cosine achieving state-of-the-art results on the SWE-bench benchmark and Distyl ranking 1st on the BIRD-SQL benchmark. This feature is available to all paid tiers, with 1M free training tokens daily until September 23. Read more: OpenAI
• Upwork Becomes an AI Powerhouse with OpenAI Integration ??: Upwork has embraced AI across its operations by integrating OpenAI’s models, improving productivity and enhancing customer experiences. Key features include the Job Post Generator, which cuts job post creation time by 80%, Upwork Chat Pro, and Uma, Upwork’s AI companion that boosts client spending. These AI-driven innovations are transforming Upwork’s platform, helping both clients and freelancers achieve better results. Read more: OpenAI
• BrightEdge Releases First-Ever Insights on OpenAI’s SearchGPT ??: BrightEdge has published an in-depth guide on OpenAI’s new AI-powered search engine, SearchGPT, providing insights into its features and comparing it with Google AI Overviews and Perplexity. The report highlights SearchGPT’s conversational approach, transparency, and use of visual aids. It also explores the potential impact of SearchGPT on search technology and its implications for industries like education, healthcare, and e-commerce. Read more: BrightEdge

? Claude Sonnet News

• Anthropic Adds LaTeX Rendering to Claude for Mathematical Precision ??: Anthropic has introduced LaTeX rendering support for its AI chatbot, Claude, allowing it to display mathematical equations and expressions consistently. This feature is available as a preview and can be enabled for new conversations on Claude's platform, making it easier to handle complex mathematical content. Learn more: Anthropic on X
• Anthropic's Claude Surpasses $1M in Mobile App Revenue ??: Anthropic's AI chatbot, Claude, has achieved a significant milestone by surpassing $1 million in revenue from its mobile app. This success highlights the growing demand for AI-powered tools and the strong market presence of Claude. The app's revenue growth is a testament to its popularity and effectiveness in providing AI-driven solutions to users. Read more: TechCrunch

?? Humanoid Robots

• Former Huawei ‘Genius Youth’ Launches Humanoid Robots to Rival Tesla’s Optimus ??: Agibot, a robotics start-up founded by former Huawei recruit Peng Zhihui, has unveiled a series of humanoid robots aimed at challenging Tesla's Optimus. The flagship model, Yuanzheng A2, is equipped with AI and sensors for complex tasks. Backed by investors like HongShan and BYD, Agibot aims to ship 300 units by year-end, positioning itself as a key player in the growing market for humanoid robots in China. Read more: SCMP
• Boston Dynamics Showcases Atlas Robot's Push-Up Skills ??: Boston Dynamics has released a new video showing its Atlas robot performing push-ups, demonstrating significant advancements in dynamic movement control. This latest display highlights the robot's improved agility and precision, underscoring the ongoing progress in robotic capabilities. Learn more: Boston Dynamics on X

• iRonCub3: World's First Jet-Powered Humanoid Unveiled ??: Researchers have introduced iRonCub3, the world's first jet-powered humanoid robot. Equipped with mini jet engines, this advanced robot can fly and perform complex aerial maneuvers. Ironcub3 represents a significant leap in robotics, combining human-like mobility with jet propulsion, potentially opening new possibilities for search and rescue missions, space exploration, and other challenging environments. Read more: Interesting Engineering

?? Corporate AI Developments

• ElevenLabs Launches Global Text-to-Speech App with 32 Language Support ??: ElevenLabs has globally released its Reader app, which allows users to convert text content into speech in 32 languages, including Spanish, French, and Hindi. The app, initially launched in the U.S., U.K., and Canada, now features hundreds of new voices, including those of famous actors. Powered by the Turbo v2.5 model, the app offers improved quality and reduced latency. ElevenLabs plans to add more features, including offline support and audio snippet sharing. Read more: TechCrunch
• Microsoft Unveils PHI-3.5 Model, Surpassing Competitors ??: Microsoft has released its new PHI-3.5 model, outperforming rivals like Google and OpenAI. This advanced model boasts significant improvements in natural language processing, setting a new standard in AI capabilities. The PHI-3.5 is designed to enhance applications across various industries, marking a major step forward in AI technology. Read more: VentureBeat
• Midjourney Updates: Web Access and New Image Editor ??: Midjourney has opened its web platform to everyone, now allowing users to explore and create through a custom-built experience. Free trials are temporarily available for new users. Additionally, a new web editor has been released, combining various image editing features like reframe, repaint, and zoom into a unified interface. The update also includes syncing between web Rooms and Discord channels, enhancing the overall user experience. Read more: Midjourney
• Andreessen Horowitz Shares List of 100 AI-Powered Apps ??: Andreessen Horowitz (a16z) has released a curated list of 100 generative AI apps, showcasing innovative applications across various industries. This collection highlights the growing impact of AI in areas such as productivity, creativity, and enterprise solutions. The list is designed to inspire entrepreneurs and developers by providing examples of successful AI implementations. Read more: a16z
• Perplexity AI to Start Running Search Ads in Q4 2024 ??: Perplexity AI, a search engine leveraging AI technology, plans to introduce search ads in the fourth quarter of 2024. This move marks the company's first venture into monetising its platform through advertising, aiming to capitalise on its growing user base. The ads will be integrated into search results, similar to traditional search engines, as Perplexity explores revenue generation opportunities. Read more: CNBC
• Salesforce Expands AI Capabilities with XGen-MM and Einstein Sales Agents ????: Salesforce has launched two significant AI advancements: XGen-MM, a set of open-source multimodal AI models, and Einstein Sales Agents. XGen-MM enhances visual-language understanding, enabling advanced tasks like image captioning. Meanwhile, Einstein Sales Agents are AI-driven tools designed to boost sales efficiency by automating routine tasks and providing intelligent recommendations. Both innovations integrate seamlessly into Salesforce’s ecosystem, pushing the boundaries of AI in business applications. Read more: VentureBeat | Salesforce.
• NVIDIA and Mistral AI Launch Minitron-8B for Smaller Language Model Applications ??: NVIDIA and Mistral AI have introduced the Minitron-8B, an 8-billion parameter language model designed for efficiency and performance in smaller-scale applications. The model leverages NVIDIA's NeMo framework to deliver robust AI capabilities while minimising computational resources, making it ideal for businesses seeking powerful AI solutions without the need for extensive infrastructure. Read more: NVIDIA Blog
• ElevenLabs Cuts Pricing for Turbo Models and Introduces Business Plan ??: ElevenLabs has reduced the cost of its Turbo v2 and v2.5 models by 50%, with each text character now consuming only 0.5 credits. They've also launched a new Business plan offering pricing as low as $50 per million text characters, making it more affordable to scale audio production. Additionally, credits can now roll over for up to two months. Learn more: ElevenLabs Pricing
• Neuralink's Second Trial Implant Shows Success, Avoids Previous Issues ??: Elon Musk's Neuralink announced that its second human trial implant has been successful, with no issues of "thread retraction" that affected the first patient. The device, designed to help paralysed individuals control digital devices with their thoughts, is now working well, allowing the second patient to play video games and design 3D objects. Neuralink addressed earlier issues by improving surgical techniques, ensuring better stability of the implant. Read more: Reuters
• Google Unveils New Android AI Feature to Lock Out Phone Thieves ??: Google has introduced a new AI-powered feature for Android that instantly locks thieves out of stolen phones. This cutting-edge security measure detects suspicious activity and immediately restricts access to the device, making it much harder for thieves to exploit stolen phones. The feature enhances the overall security of Android devices, offering users better protection against theft. Read more: Forbes

?? AI Innovations

• Deep-Live-Cam: Real-Time Face Swapping Tool ????: Deep-Live-Cam is a GitHub project that enables real-time face swapping and the creation of video deepfakes using only a single image. The tool is designed to support artists in animating characters or modeling clothing, with a strong emphasis on ethical use. Safeguards are in place to prevent inappropriate content, and users are encouraged to obtain consent and disclose when content is a deepfake. The setup requires Python and additional dependencies, with support for GPU acceleration on various platforms. Users can customise the process via command line options, making it a versatile tool for creative AI-generated media. Learn more: GitHub | See examples on X.
• Hotshot Launches AI-Powered Text-to-Video Generator ??: Hotshot has introduced a new AI tool that transforms text into video, revolutionising content creation. This generator enables users to input text prompts and receive fully-produced videos, streamlining the video production process. Aimed at marketers, educators, and content creators, this tool offers a fast and efficient way to create visual content without needing advanced editing skills, setting a new standard in video creation. Read more: VentureBeat
• D-ID Launches AI Video Translation Tool with Voice Cloning and Lip Sync ??: D-ID has introduced an AI-powered video translation tool that includes advanced features like voice cloning and lip-syncing. This tool allows users to translate videos into multiple languages while maintaining the original speaker's voice and synchronising their lip movements with the translated audio. It's designed to enhance accessibility and communication across different languages, particularly for global content creators and businesses. Read more: TechCrunch
• Ideogram Launches Version 2.0: Enhanced AI-Powered Text-to-Image Generation ??: Ideogram has unveiled its Version 2.0, which brings advanced capabilities in AI-powered text-to-image generation. The new version offers improved image quality, more customisation options, and better integration with various workflows. It’s designed to cater to artists, designers, and businesses looking to create unique visual content quickly and efficiently. Read more: Ideogram
• NVIDIA Develops AI Model for Predicting Extreme Weather ???: NVIDIA has introduced an AI-powered weather model capable of predicting extreme weather events with greater accuracy. The model, designed to anticipate phenomena like hurricanes and heatwaves, aims to enhance climate predictions and provide crucial data for mitigating the impacts of climate change. This innovation represents a significant advancement in the use of AI for understanding and responding to climate-related challenges. Read more: Axios
• Meshy 4 Brings Sci-Fi-Level AI to 3D Modelling ??: Meshy 4, a new AI-driven tool, is revolutionising 3D modelling and design with advanced features that streamline the creation of highly detailed models. Leveraging cutting-edge AI, Meshy 4 automates complex tasks, making the design process more efficient and accessible. This tool is set to significantly impact industries like gaming, architecture, and animation by enabling designers to achieve sci-fi-level precision and creativity. Read more: VentureBeat

?? AI in Business

• Best Buy Introduces AI-Powered Delivery Tracking, Shaping the Future of Retail Logistics ??: Best Buy has launched an AI-powered delivery tracking system, marking a significant shift in retail logistics. This new system enhances the customer experience by providing real-time updates and more accurate delivery windows. The move highlights the increasing role of AI in optimising supply chain operations, ensuring timely and efficient deliveries. Read more: PYMNTS
• Google Develops AI Model for Cough Disease Detection ??: Google has announced an AI model capable of detecting diseases by analysing the sound of a cough. This innovative approach could help identify conditions like respiratory infections and other related illnesses by leveraging AI to assess audio patterns. The technology represents a significant step forward in non-invasive health diagnostics, with potential applications in early disease detection and remote healthcare. Read more: Google Blog
• AI Predicts Tipping Points for Future Disasters ??: Scientists have developed an AI algorithm capable of predicting critical "tipping points" in systems like ecosystems and financial markets. By analysing patterns in data, the AI aims to forecast catastrophic events such as ecological collapse or pandemics, providing early warnings that could help mitigate damage. The AI was tested on real-world data, accurately predicting the transformation of tropical forests into savannah, and researchers plan to expand its use to other complex systems. Read more: Live Science
• Epic Systems Announces Over 100 New AI Features for Healthcare ??: Epic Systems is developing more than 100 AI-powered features aimed at enhancing both doctor and patient experiences. Upcoming tools include generative AI for simplifying medical communications, automating prescription orders, and calculating wound measurements from images. Epic also teased future capabilities, such as AI-driven patient recovery assessments through MyChart. These innovations mark a significant step in integrating AI into healthcare operations. Read more: CNBC
• SleekFlow Secures $7M to Expand AI-Powered Social Commerce Across Asia ??: SleekFlow, a social commerce platform based in Singapore and Hong Kong, has raised $7 million to enhance its AI capabilities and expand across Southeast Asia, the Middle East, and Europe. The platform specialises in conversational AI for customer engagement, offering omnichannel marketing, in-chat payments, and CRM integration. With a focus on streamlining customer interactions, SleekFlow aims to capitalise on the growing social commerce market in Asia. Read more: TechCrunch
• USAID Adopts ChatGPT Enterprise for Enhanced Operations ???: The U.S. Agency for International Development (USAID) is integrating OpenAI’s ChatGPT Enterprise into its operations to improve efficiency and decision-making. This adoption is part of a broader effort to leverage AI tools across government agencies. ChatGPT Enterprise will assist in various tasks, from automating routine processes to providing data-driven insights, thereby enhancing the agency's overall capabilities. Read more: FedScoop
• GPT-4 Surprises Scientists with Protein Modeling Capabilities ??: OpenAI's GPT-4 has unexpectedly demonstrated the ability to model basic protein structures, a task typically reserved for specialised AI systems. This discovery highlights GPT-4's versatility and potential in scientific research, particularly in areas like biology where understanding protein structures is crucial for drug discovery and other applications. The findings suggest that general-purpose AI models like GPT-4 could play a significant role in advancing scientific understanding in unexpected ways. Read more: The Decoder
• Amazon Cloud CEO Predicts AI Will Transform Software Engineering ??: AWS CEO Matt Garman predicts that within the next 24 months, most software engineers may no longer need to code, as AI will handle much of the coding work. In a leaked discussion, Garman emphasised that this shift will change the role of developers, who will need to focus more on understanding customer needs and the end product rather than writing code. This transformation could significantly alter the software engineering landscape. Read more: Entrepreneur
• AI as a Game Changer for Accessibility ??: AI, including ChatGPT, is revolutionising accessibility for people with disabilities by enhancing communication, simplifying tasks, and improving access to information. ChatGPT can assist with voice-to-text and text-to-speech for visual or motor impairments, automate repetitive tasks, and provide personalised support for learning disabilities. These capabilities empower users, making technology more inclusive. Read more: MIT Technology Review

? Responsible AI

• McAfee Integrates Deepfake Detector in Lenovo's New Copilot Plus PCs ???: McAfee has launched a deepfake detection tool integrated into Lenovo's Copilot Plus PCs. This new feature enhances security by identifying AI-generated fake content, providing users with a safeguard against potential misinformation. The integration signifies a growing trend in the tech industry to combat deepfake threats, ensuring a more secure digital environment for users. Read more: CNET
• Pulitzer-Winning Authors Join Copyright Lawsuit Against OpenAI and Microsoft ??Prominent authors, including Pulitzer Prize winners, have filed a class-action lawsuit against OpenAI and Microsoft, alleging that their copyrighted works were used without permission to train AI models. The lawsuit, backed by the Authors Guild, claims that this infringement harms the authors' livelihoods and seeks damages along with an injunction to prevent further use of their works. This legal challenge underscores the growing tension between AI companies and content creators over the use of copyrighted material. Read more: Reuters | Authors Guild
• Microsoft Pauses Windows AI Feature Testing Due to Recall Concerns ???: Microsoft has decided to halt testing of a new AI feature in Windows set for release in October, citing potential issues that may require a recall. The pause aims to address concerns before wider rollout, ensuring the feature meets quality standards. This decision underscores Microsoft's cautious approach to integrating AI in its core products. Read more: The Verge
• Hong Kong Monetary Authority Issues Generative AI Guidelines ??: The Hong Kong Monetary Authority (HKMA) has released new guidelines on the use of generative AI in the financial sector. These guidelines focus on ensuring that AI technologies are used responsibly, with an emphasis on data privacy, security, and ethical considerations. The HKMA aims to support innovation while safeguarding the integrity and stability of the financial system. Read more: Cointelegraph
• OpenAI Warns California AI Bill Could Stifle Innovation ?? OpenAI has raised concerns that a proposed AI regulation bill in California may hinder innovation in the U.S. The bill aims to regulate AI development and use, but OpenAI argues that the legislation could impose overly restrictive measures, potentially slowing down AI advancements and the competitive edge of U.S. companies in the global market. Read more: AI News
• U.S. Government and Humane Intelligence Launch AI Flaw-Hunting Contest ??: The U.S. National Institute of Standards and Technology (NIST) and the nonprofit Humane Intelligence are inviting U.S. residents to participate in a nationwide "red-teaming" contest to find flaws in generative AI systems. This initiative aims to democratise AI evaluation, involving developers and the public in assessing AI models for security, resilience, and ethical integrity. The final event will occur at the Conference on Applied Machine Learning in Information Security (CAMLIS) in Virginia. Read more: Wired

?? Videos to Watch

AlphaGo the Movie by DeepMind - A Landmark Achievement in AI ????

"AlphaGo" is a documentary that captures a significant moment in AI history: DeepMind's AI system defeating a world champion at the ancient and complex game of Go.

Key points:

1. Go Complexity: Go is exponentially more complex than chess, with approximately 10^170 possible board configurations which is more than atoms in the universe. This complexity long posed a significant challenge for AI researchers. Traditional AI methods that worked well for games like chess were not sufficient for Go, requiring new approaches in machine learning and AI development.
2. AI Creativity - Move 37: The documentary highlights AI's creative potential, especially through the famous "Move 37". This move, played in the second game against Lee Sedol, was described by experts as "beautiful" and unlike any human move. It demonstrated AlphaGo's ability to devise novel, effective strategies beyond human convention.
3. Human Element: The documentary captures the emotions and reactions of both the DeepMind team and the Go champions, adding a personal dimension to this technological achievement.
4. Development Insights: Viewers get a glimpse into the process of developing cutting-edge AI systems, offering valuable perspective for those in the field.
5. Broader Implications: The film explores how AI advancements might reshape human creativity, problem-solving, and various aspects of our future.

"AlphaGo" is more than just a record of a game; it's a window into the potential of AI and its impact on human expertise. Whether you're an AI professional or simply interested in technology's future, this documentary provides an engaging and thought-provoking viewing experience.

Watch it here: AlphaGo - The Movie | Full Documentary

After viewing, consider: How might AI's problem-solving abilities be applied to other complex challenges in our world?

??? We'd Love to Hear from You!

Did you find this edition insightful? Like, share your thoughts, and ask us anything.

Stay curious, and until next time, enjoy exploring the vast possibilities AI offers.

???? Subscribe on LinkedIn The ChatGPT Observer Newsletter

As we explore the captivating world of AI technologies, it's important to note that the content of this article is a product of human effort, enhanced by AI tools and additional resources. We have made sure to properly attribute and link all referenced articles to their original sources.