AI Legal/Regulatory headwinds increase: The EU expanded its focus on both Microsoft and Apple for different perceived violations. With Microsoft it was on antitrust violations over its Teams software applications. Apple is facing a non-compliance investigation under the European Digital Markets Act (DMA). Apple of course already separately signaled AI Feature delays in the EU, specifically with ‘Apple Intelligence’, its version of AI. The EU Competition Commissioner Margrethe Vestager responded that this Apple actions ‘shows anticompetitive behavior’. Separately, the EU is exploring an antitrust probe into Microsoft’s OpenAI deal. Back in the US, the RIAA sued two leading AI music startups Udio and Suno, while reports indicated Google YouTube may be in negotiations on AI music licensing deals with some major music labels. Then, Microsoft AI head Suleyman says web content is ‘freeware’ under ‘Fair Use’. Separately, OpenAI delays its long-anticipated AI Voice technology a little bit longer. These technologies have been facing heightened regulatory focus, especially in an election year. OpenAI continues to have a rich product pipeline for the rest of the year. A broader look at US vs US Big Tech actions here.
AI Conversation Chatbots accelerate: Both Google and Meta are expanding their efforts with personalized, conversational AI chatbots. This while Character.ai is expanding multimodal Voice access to its popular AI chatbots. This while other AI chatbot and ‘companion’ companies like Replika, Kindroid and others also expand their offerings in the ‘smart agent/companion’ segment of customizable conversational AI chatbots. More here.
Media use of AI “Digital Twins” expands: NBC announced an AI version of legendary sportscaster Al Michaels in the Paris Olympics, underlining the increased interest in ‘Digital Twins’ in traditional media and hollywood. In the meantime, big tech companies rush to update their Terms of Service for more AI data training and LLM applications. Combined with the aforementioned RIAA lawsuits against AI music startups and Google YouTube negotiations with music labels, the tug-of-war over digital rights for AI and digital twins applications continue. More here.
Character.ai balances LLM AI costs vs quality: This week also saw an interesting case study on leading AI companies striving to increase Ai cloud compute efficiencies, while delivering consistent high-quality services to millions of users. Case in point was Character.ai with its millions of conversational, customizable AI chatbots seeing rising customer complaints on quality. The company provided detailed explanations on impressive technical tweaks to their AI compute architecture that increased cost and performance efficiencies, while serving up to 20% of Google’s Search volume per day. More broader context of this ‘quality vs efficiencies’ trend on LLM AI companies here.
Apple expands external LLM AI partner list: Apple continues to potentially expand its list of external LLM AI partners beyond OpenAI. After dropping hints of conversations with Google and Anthropic, it now looks like Meta with its Meta AI LLM technology may also be in the mix. Apple of course continues to benefit from its close business relationship with Google in a $20 billion/year deal for exclusive Google Search on Apple devices. There are potentially several advantages for Apple to consider multiple external LLM AI partners, which I go into more detail here. Apple remains in a relatively unique position to leverage its distribution capabilities for a wide range of LLM AI companies.
Other AI Readings for weekend:
OpenAI rolls out GPT-4 finding GPT-4 mistakes and errorsr.
Meta expands open-source LLM Compiler tools, potentially accelerating developers AI workload, and reiterates Open source LLM AI support.
Thanks for joining this Saturday with your beverage of choice.
Up next, the Sunday ‘The Bigger Picture’ tomorrow. Stay tuned.
(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)