?? The AI & Data Digest - Edition 6 ??

Hi everyone and welcome back to the AI & Data Digest. This week's edition is packed with the latest insights, news, and expert opinions on AI, data, and digital transformation. Let's dive straight into your essential roundup of this week's top news insights and opportunities.

?? Welcome to Your Weekly Intelligence Brief!

?? Top 10 News Beats This Week:

1. ?? OpenAI Insiders Warn of 'Reckless' Culture: Current and former employees raise alarms about OpenAI's pursuit of AGI, prioritising profit over safety. An open letter calls for transparency and whistle-blower protections. [NYT]

1. ?? Apple Introduces 'Apple Intelligence' for Seamless AI Integration: Apple's new AI system integrates on-device and cloud capabilities for enhanced privacy and functionality. [Bloomberg]

1. ???? Salesforce Launches AI Centre in London: Salesforce's new AI centre in London aims to create 500,000 AI-related jobs and foster industry collaboration. [CNBC]

1. ??? Microsoft Adjusts Controversial Screenshot Feature: Microsoft changes its "Recall" feature to opt-in after privacy concerns, enhancing safeguards ahead of Copilot+ PCs launch. [BBC]

1. ?? FTC Investigates Microsoft's Deal with AI Startup Inflection: The FTC is scrutinising Microsoft's $650M deal with Inflection, part of broader AI industry regulation efforts. [CNN]

1. ?? AWS Invests in Tools for Managing Large Language Models: AWS is enhancing SageMaker with new tools for operationalising and managing LLMs, addressing the rising demand for generative AI. [InfoWorld]

1. ??? AI Designs Trophy for Formula 1 AWS Grand Prix du Canada: AWS used generative AI to design the first F1 trophy, blending AI innovation with traditional craftsmanship. [Amazon]

1. ?? U.S. Regulators Target Nvidia, Microsoft, and OpenAI in Antitrust Probes: The DOJ and FTC are investigating potential antitrust violations by major AI players like Nvidia, Microsoft, and OpenAI. [NYT]

1. ?? Meta's AI Training Plans Face Criticism: Digital rights groups criticise Meta's use of public posts and images for AI training, calling it an "abuse of personal data." [BBC]

1. ?? Snowflake Faces Data Theft Claims Amid MFA Policy Scrutiny: Advance Auto Parts is the latest victim in a series of data thefts linked to lenient Snowflake MFA policies. [TechCrunch]

?? My Insights:

This week, I delved into the complexities of AI and data strategies. In my latest blog, "50+ Questions to Build Your AI Strategy Around". I discuss how to frame your AI initiatives with the right questions, ensuring alignment with business goals and maximising value.

Another highlight from my work is exploring "Navigating AI Risks with Key Risk Objectives and Indicators", where I outline strategies for identifying and managing risks in AI projects. This piece is essential for organisations aiming to deploy AI responsibly and effectively.

?? Guest Blog:

This week's guest blog is by Fanghua Yu, Principal Architect at Neo4j, discussing "Data Profiling: A Holistic View of Data using Neo4j." Fanghua emphasises that data quality is the key to driving trustworthy AI and suggests leveraging Knowledge Graphs to uplift data quality where possible. He addresses the common issue of data silos, often resulting from sub-standard data governance controls and practices. By improving these practices, organisations can better handle AI, especially in managing generative AI's hallucinations. Fanghua's insights on leveraging graph databases for better data analysis are a must-read for data professionals. Read the blog here.

?? Success Stories:

1. Canva's AI-Powered Magic Studio: Canva’s AI-powered Magic Studio has been used 5 billion times, transforming the creative process with tools powered by OpenAI’s GPT-4, including Magic Write, Magic Design, and Magic Switch. Read more

1. Volvo Group's AI-Driven Invoice Processing: Volvo Group streamlined invoice and claims processing with Microsoft Azure AI, saving over 10,000 manual hours and improving employee satisfaction. Read more

1. Capital One Enhances Fraud Protection with AI: Capital One uses machine learning on AWS to protect customers while reducing false positives in fraud detection, enhancing user experience and operational efficiency. Read more

?? Upcoming AI Events in London This June:

1. ?? Artificial Intelligence for Cybersecurity -? 12th-13th June, Tobacco Dock, London?

1. ?? The AI Summit -? 12th-13th June, Tobacco Dock, London?

1. ?? The AI Summit - Afterparty Sponsored by HPE & NVIDIA -? 12th June, Tobacco Dock, London?

1. ?? AI Hardware & Edge Summit -? 18th-19th June, ETC Venues, Aldersgate Street, St.Pauls, London

1. ?? Moving Target: AI Risk Management Fundamentals -? 25 Copthall Avenue London EC2R 7BP

?? Research & Reports:

This week's featured report is "Understanding AI Risks: A Comprehensive Framework" by Yann LeCun. This seminal paper from 2015 explores the foundational concepts of AI risk, providing a framework that remains highly relevant today. Key highlights include the differentiation between short-term and long-term risks, the importance of robustness and interpretability in AI models, and the necessity of establishing clear ethical guidelines. One particularly interesting section discusses the need for continuous monitoring and assessment of AI systems to mitigate unforeseen consequences. LeCun's work underscores the importance of balancing innovation with safety to ensure AI technologies are developed responsibly. Read the paper here.?

?? Regulatory & Ethics Watch:

???? California AI Protection Act: A proposed bill in California, introduced by State Senator Scott Wiener, aims to establish "common sense safety standards" for AI companies. The legislation requires companies to take steps to prevent their models from causing harm, ensure AI systems can be shut down, and disclose compliance efforts to a newly established "Frontier Model Division" within the California Department of Technology. The bill has received support from AI pioneers like Geoffrey Hinton and Yoshua Bengio but faces criticism from the tech community, which fears it could stifle innovation and place an undue burden on developers. If passed, this legislation could set a precedent for AI regulation across the U.S., highlighting the balance needed between innovation and safety.

?? Introducing For Humanity: Earlier this week, I met with Ryan Carrier from ForHumanity, a non profit organisation dedicated to examining and analysing the downside risks associated with the ubiquitous advance of AI & Automation. Their mission is to engage in risk mitigation and ensure the optimal outcome, fostering a community of around 2100 people from 97 countries who work together to establish best practice guidance, controls, and standards for the safe adoption of AI across the world.

Meeting with Ryan was brilliant and the beliefs exhibited by ForHumanity really resonated with me. It is an organisation that I would really like to do more with in the future. After a short cursory glance across the website, I found this piece edited by Enrico Panai really helpful.

The "Code of Data Ethics" (CoDE), edited by Enrico Panai, is a comprehensive document aimed at guiding organisations in making ethical decisions related to data and AI. The CoDE is inspired by Luciano Floridi's works and aims to provide an operational approach to data ethics. It consists of a Public Part, which includes general principles and methodologies, and a Private Part, detailing values and approaches specific to the organisation.

Key Highlights from the document include:

1. The Importance of Data Ethics: The CoDE emphasises the necessity of adopting a Code of Data Ethics in mature information societies. It serves to guide data stewards and custodians in ethical decision-making processes, aligning with the organisation's broader ethical framework.

2. Ethical Framework: The document outlines an ethical framework based on information ethics, focusing on the well-being and prosperity of the digital environment. It promotes moral evaluations that align with the organisation's principles, emphasising beneficence, non-maleficence, autonomy, justice, and explicability.

3. Distributed Moral Responsibility: A significant concept introduced is the idea of distributed morality. In complex AI systems, responsibility for ethical outcomes is shared among all agents involved. This approach aims to mitigate ethical risks by ensuring that all participants are aware of their moral responsibilities, thereby preventing negative consequences.

4. Role of the Data Ethicist: The document highlights the crucial role of Data Ethicists or Chief Data Ethics Officers. These professionals are responsible for maintaining ethical standards, supporting developers, and ensuring that ethical considerations are integrated into all stages of data processing and AI system development.

5. Levels of Abstraction (LoA): The method of levels of abstraction is crucial for ethical analysis. It allows for the appropriate framing of moral situations at different levels of granularity, ensuring that ethical evaluations are comprehensive and contextually relevant.

6. Semantic Continuity: Maintaining the semantic capital, or the meaningfulness of data, is emphasised. Ethical actions should aim at preserving the integrity and context of data to ensure it remains useful and reliable over time.

The CoDE provides a robust framework for integrating ethics into data management and AI systems, highlighting the importance of a well-defined ethical structure within organisations. It is a valuable resource for anyone looking to understand and implement data ethics in a practical, effective manner.

I highly recommend checking out ForHumanity's website for more insightful resources and guidelines on ethical AI and data practices.

?? Careers & Opportunities:

Explore these exciting AI and data opportunities in the UK:

1. Data Scientist Manager, Mastercard - Fintech sector, London. Apply here ????
2. Senior Data Scientist, Field - Trading sector, London. Apply here ??
3. Senior Machine Learning Engineer, OVO - AI sector, ￡64-92k, London. Apply here ??
4. Data Scientist, Capital on Tap - Fintech sector, London. Apply here ??
5. Senior Data Scientist, Monzo - Fintech sector, London. Apply here ??

Thanks for checking in!

?? Connect With Me! Engage with me further on these topics and more by connecting on LinkedIn.

?? Like & Share Enjoyed this newsletter? Hit like and share it within your network to help others stay on top of the latest in AI and data! If you have any feedback or suggestions, let me know! Thanks for reading!

Thank you,

Ben @ The AI & Data Digest Team