Advisory Board Member, Lighthouse Canton & Portfolio Manager for LC GenInnov Global Innovation Fund
The tech world is abuzz with GPT-4o. But like the shiny new iPad or AlphaFold 3.0 last week, this excitement will surely fade in hours. What won’t fade are the seismic implications this “small” model has set in motion.
Let’s be clear: GPT-4o isn’t just another LLM. In some ways, this is about a fundamental shift in how we interact with AI. Under the hood, how it operates has significant, novel implications for the entire hardware chain, with new winners and relative losers. Elsewhere for regulators and parents, for educators or corporates, for the silence lovers and the gregarious, the life-changing ripples will have implications that we will not stop talking for years, even if the eventual model to dominate the use-case turns out to be something other than GPT 4o (expect every model to have the same features in weeks and more).
Small is more beautiful and workable:
The arrival of smaller, more efficient models like GPT-4o throws a wrench in the prevailing narrative of “bigger is better”. This has enormous consequences:
GPU Demand in Question: The assumption that massive cloud servers with high-end GPUs are the future of AI inference takes a hit relative to otherwise.
Cloud Revenue at Risk: If processing moves to personal devices, cloud giants like Amazon and Azure could see a dent in their long-term growth expectations compared to what is assumed now.
The Rise of End-Point AI: Expect a surge in demand for specialized chips optimized for on-device AI processing – think TPUs, FPGAs, and custom silicon.
The Death of the Chatbox:
GPT-4o’s multimodality signals the end of an era. Chatbots are so 2022. The future is immersive, interactive, and goes way beyond text.
Multimodality Becomes the Norm: Expect to see this feature in almost every new LLM, opening up a universe of use cases.
Consumer Hardware Gets a Kickstart: We’re talking cameras, sensors, and new input methods – all designed to take advantage of multimodal AI.
The Apple of OpenAI’s Eye:
The launch on Mac systems, the rumored partnership – the clues point to Apple becoming OpenAI’s new BFF. This has major consequences:
Google’s Grip Loosens: A potential OpenAI-Apple alliance threatens Google’s long-standing dominance in search, particularly on Apple devices.
Microsoft Gets Sidelined?: This weakens the Microsoft-OpenAI relationship, explaining Microsoft’s recent push for its own in-house models.
The Search Wars Heat Up: Expect Google to double down on AI innovation and monetization to counter this threat.
The Future is now, with personalization becoming the norm in the quarters ahead, compute moving to consumer pockets, and the near-doubling of entities with human voice!
This article is contributed by Nilesh Jasani, Advisory Board Member, Lighthouse Canton and Portfolio Manager for LC GenInnov Global Innovation Fund.
.jpg)
.jpg)
.jpg)
.jpg)