Bruno R.’s Post

I think this is the best article i've read about AI. "...generative AI, and ChatGPT itself, is a commodity technology that is most useful when it is: Embedded in a system that gives it broader context about the user (which might be search, social, a device OS, or a vertical application) and Unbundled into individual features (ditto), which are inherently easier to run as small power-efficient models on small power-efficient devices on the edge (paid for by users, not your capex budget) - which is just as well, because… This stuff will never work for the mass-market if we have marginal cost every time the user presses ‘OK’ and we need a fleet of new nuclear power-stations to power it all."

Thought provoking article from Benedict Evans, which makes a strong case that LLMs will eventually become commodity infrastructure. With at least 3 players (Google, Anthropic, OpenAI) now developing top-of-the-line frontier models (excluding open source) and new ones joining the race, it is hard to argue with this perspective. The big question for me is whether and how these players can build sufficiently strong ecosystems as well as consumer and enterprise value propositions that will enable them to lock in customers (just as Apple have done with theirs).

Is everyone heading in the wrong direction with GenAI? 🧐 If we look at how Apple presented their version of GenAI. The answer could be: yes. Below an interesting analysis by Benedikt Evans. 💡 Take aways: - Apple is treating LLMs as a commodity. - Apple's AI focuses on user context and integrating features into devices. They separate between a world model (Open AI, others) and a context model (their own) - Instead of general-purpose chatbots, Apple uses AI to enhance specific tasks. In the Apple world, there won't be prompts. - This approach minimizes errors and maximizes efficiency. Is Apple setting a new standard for AI? One thing is for sure: It's an interesting time to be in the tech space. Happy Friday!☀️

Great article from Benedict Evans looking at Apple's approach to LLMs and what it means for the future and impact of Generative AI. Apple's approach gives their answer to the fundamental question "is this a new general purpose tool, where one product from one company does the work of hundreds of pieces of software from hundreds of companies, or is this a generic technology that will enable features inside products from hundreds or thousands of companies?" Apple thinks that LLM will be a generic commodity technology which only becomes interesting when when it is integrated into other applications and systems which give it the context to provide useful and trustworthy outputs. I'm inclined to agree. In most cases that matter, sounding generically like a human response is not sufficient to be valuable, it needs to be a response that understands the context. A good doctor doesn't just say doctorish things. A good doctor provides an expert opinion based on specific data and context, combined with general experience. The result, incidentally, sounds doctorish.

Great article from Benedict Evans on Apple's move into AI, and what it might mean for the space. Interesting to see more thinking that GenAI might very well be a commodity - a technology - rather than a product on its own. And, as the article points out, if that's true, then the only true moat for the AI sector is capital. It's going to be interesting to see if this trend continues or if one of the players manages to change the dynamic, but for now, they're all just playing leapfrog with LLMs. https://lnkd.in/ebF8mmKs

Must-read #GenAI analysis from the brilliant Benedict Evans addressing "the single core question about the future of Generative AI - is this a new general purpose tool, where one product from one company does the work of hundreds of pieces of software from hundreds of companies, or is this a generic technology that will enable features inside products from hundreds or thousands of companies?" Evans shows how Apple's strategy reveals they believe the latter, with profound implications. At ModuleQ, our experience deploying #AI in global financial institutions has led us to the same conclusion: personal context is key, and #LLM tech is often most effective behind the scenes to enable focused, high-value use cases. #UnpromptedAI https://lnkd.in/g-t92rup

Pretty interesting and well presented analysis explaining #Apple Intelligence and #AI with their #strategy versus the wider angle approach.

Apple finally showed their AI hand. Their first formal reveal demonstrated a set of modest but interesting #GenAI-enabled features for iPhone as well as some hope-for-the-best improvements to Siri. Their conservative toe in the water could be a bellwether for a broader tech industry recognition of the economic reality of AI hype. But that recognition does not necessarily mean that AI won't be transformative. It could be, instead, a welcome step toward more pragmatic thinking about where lasting AI-enabled transformation can really occur. Apple famously lets others storm the beach and shows up later with more refined consumer experiences based on lessons learned by others. While there were no bold surprises, there were some novel technology approaches as well as some confirmations. Apple introduced their notion of "semantic index" and "private cloud compute." The first is Apple's version of capturing all of our activity on our devices and then making that available to us in an AI-enhanced way (nevermind cynics would say they've been doing that all along, just for their use.) And, they are building ways to store our AI-enhanced data in the cloud (away from our device) that aspire to protect privacy and deter malicious activity. Also, Apple's announcements implicitly confirmed they have no intention of trying to compete with the user-facing, frontier models of OpenAI and Anthropic. All of that says they are making their AI bet on the "mundane utility" side of the AI equation rather than the more immersive "thought partner" experience of ChatGPT and Claude. Many will point to this Apple strategy in particular, and the whole industry tilt in general, as acknowledgement that the only way for Big Tech to get any positive ROI on the massive dollar investments required to build and field frontier AI technology is to peanut butter AI into existing consumer experiences. And, that may well be true. But, it ignores the slower, perhaps less lucrative, but potentially more impactful scenario where AI transformation happens one individual, one classroom, one enterprise at a time, rather than a Big Bang upheaval. The problem for the Apple's of the world is that this one-human-at-a-time adoption of the technology doesn't make headlines or spin straw into gold. That is mostly because of transactional friction--a fancy way to say, use of AI as a "thought partner" requires work on the part of humans. The good news is these micro-transformations can result in many rising boats of human capital and opportunity that benefit every day people in material ways and, in turn, aggregate into the kind of transformation that benefits even poor little rich boy Apple.

For Apple, AI Is Personal

On GenerativeAI, "is this a new general purpose tool, where one product from one company does the work of hundreds of pieces of software from hundreds of companies, or is this a generic technology that will enable features inside products from hundreds or thousands of companies?" I think it could be both as it speaks to GenAI's assistive and contextual capabilities, though the emergence of a general purpose tool across domains may take a while compared to having in-context AI embedded in features of different products. #genai #ai #product #productmanagement https://lnkd.in/g5vy38h8