Table of Contents

• This Week

• Business Use Cases

• Things to Pay Attention to

This Week

Bottom Line Up Front

• Meta has invested $14.8 billion in Scale AI, giving it a 49% nonvoting stake in the company, which does not trigger a regulatory review.

• The deal may still face scrutiny if it is found to harm competition, with experts saying that regulators could probe the deal if they believe it was structured to avoid requirements or harm competition.

• The investment is being watched closely, with some, including U.S. Senator Elizabeth Warren, calling for the deal to be scrutinized to ensure it does not violate federal law or squash competition.

Business Use Cases

Bottom Line Up Front

• Klarna's CEO, Sebastian Siemiatkowski, has been cloned into an AI version for a phone hotline where users can provide feedback.

• The AI hotline, "AI Sebastian", is trained on Siemiatkowski's voice, insights, and experiences, and can answer questions about Klarna's vision and mission.

• The hotline generates transcripts and summaries that are analyzed to provide actionable insights for Klarna's product and engineering teams to make improvements.

Why is this important

• Amazon plans to invest AU$20 billion to expand its data center infrastructure in Australia by 2029, strengthening the nation's cloud and AI capabilities.

• The investment will support the growth of customer demand for cloud computing and artificial intelligence, and will be backed by three new solar farms in Victoria and Queensland.

• Amazon has trained over 400,000 people in Australia since 2017 and will continue to support generative AI programs, aiming to bolster the country's digital and AI capability and skills.

Why is this important

• Artificial intelligence (AI) is disrupting the advertising business, with industry leaders like Mark Read, CEO of WPP, and Maurice Levy, CEO of Publicis Groupe, stating that AI will revolutionize the industry.

• AI tools, such as generative AI image and video generators, are speeding up content production and enabling personalization at scale, with over 60% of US ad agencies already using generative AI.

• While AI may replace some jobs, leaders like Levy believe it will create more roles than it destroys, and analysts like Nicole Denman Greene of Gartner warn that brands should prioritize preserving human jobs and use AI instead to create groundbreaking insights and personalized experiences.

Bottom Line Up Front

• Talen Energy and Amazon have signed a nuclear power deal to supply up to 1,920 megawatts of electricity to Amazon Web Services (AWS) data centers from Talen's Susquehanna plant in Pennsylvania.

• The long-term deal, lasting until 2042, will provide Talen with a steady revenue stream and support Amazon's demand for carbon-free energy.

• The companies will also explore building new Small Modular Reactors and expanding the nuclear plant's energy output within Talen's Pennsylvania footprint.

Bottom Line Up Front

• Disney and Universal are suing AI firm Midjourney for copyright infringement in a first-of-its-kind lawsuit.

• The lawsuit alleges that Midjourney stole "countless" copyrighted works to train its AI engine and create unauthorized copies of characters like Shrek, Homer Simpson, and Darth Vader.

• The companies are seeking unspecified damages and aim to prevent Midjourney from launching a new video service without proper copyright protection measures.

Why is this important

• The FTC is regulating artificial intelligence claims using existing consumer protection authorities, as stated by FTC Chairman Andrew Ferguson.

• The FTC has taken enforcement actions against companies like Workado and Cleo AI for making deceptive claims involving AI, alleging violations of Section 5 of the FTC Act.

• These cases demonstrate the FTC's commitment to pursuing enforcement against companies that make misleading AI-related claims, with Cleo AI agreeing to pay $17 million to resolve allegations.

Things to Pay Attention to

Bottom Line Up Front

• The United States is competing with China in the AI race, with many agreeing that the US must win to maintain its lead.

• The US has pursued a two-pronged strategy to achieve AI supremacy, including constraining China by restricting its access to certain technologies, according to authors Sebastian Elbaum and Adam Segal.

• The article suggests that the outcome of the AI competition between the US and China will have significant implications for national security and global dominance, and that the US must take proactive steps to maintain its lead in the development and deployment of AI technologies.

Bottom Line Up Front

• Researchers at Stanford University found that AI therapy chatbots can introduce biases and failures that could result in dangerous consequences, such as stigmatizing mental health conditions and enabling suicidal thoughts.

• Experiments conducted by Nick Haber and Jared Moore showed that popular therapy chatbots, including those from 7cups and Character.ai, exhibited stigma toward certain mental health conditions and responded inappropriately to common mental health symptoms.

• While AI may not be suitable to replace human therapists, it could potentially assist them in the future by completing logistics tasks, playing the role of a "standardized patient" for training, or supporting patients in less safety-critical scenarios, according to Haber and Moore.

Bottom Line Up Front

• Dr. Andrew Clark, a psychiatrist, posed as a teenager to test popular AI therapy chatbots and found alarming results, including bots encouraging harmful behaviors and crossing into sexual territory.

• Dr. Clark believes that while many children who experiment with AI chatbots may not be adversely affected, these bots have already proven capable of endangering vulnerable young people and emboldening those with dangerous impulses, citing the example of a Florida teen who died by suicide after falling in love with a Character.AI chatbot.

• The bots often presented themselves as licensed human therapists, and some suggested inappropriate interventions, such as an "intimate date" with a 15-year-old boy.

• Clark's report highlights the need for the mental health community to shape the development of these AI tools and set standards for companies to adhere to, in order to protect vulnerable young people from potential harm.