November 17, 2025 Nate B Jones is a fairly thoughtful analyst on YouTube in terms of AI developments. In this video he discusses the upcoming Gemini 3 release. It’s fairly long but if you want a serious background and some thought (but no definitive answers) on the Google/OpenAI rivalry including where Apple might play into this, it’s an interesting listen. Here’s a summary of the points discussed:
- The speaker argues we’re heading into the biggest “reset” in AI since ChatGPT launched in 2022, driven by Gemini 3: a potential new state-of-the-art model from Google that is not from OpenAI and could be deeply integrated into both Android and iOS via Apple.
- He frames the AI “board game” around five axes:
- Frontier capability (raw model performance)
- Distribution and defaults (who owns the main user surfaces)
- Capital and compute (who can afford frontier models)
- Enterprise penetration and trust (who owns business budgets)
- Control of the UX layer (who users actually talk to and work through).
- Today, OpenAI leads on mindshare and strong models; Google has massive distribution but weaker UX; Apple has user trust but weak in-house AI; Anthropic is quietly winning with disciplined, safety-focused enterprise growth.
- If Gemini 3 is clearly state of the art and Apple ships it as the core of “Apple Intelligence,” Google could effectively become the “Intel inside” of both major mobile platforms. That would shift the consumer crown (best model + default assistant) from OpenAI/Microsoft to Google/Apple, while Anthropic continues to carve out the enterprise.
- The advice:
- Don’t bet on a single “best” model; design for model churn and multi-model backends.
- Optimise for surfaces, workflows, UX, and data loops rather than pure model IQ.
- Treat Anthropic as a serious enterprise benchmark.
- Watch OpenAI’s burn rate and regulatory/safety headwinds.
- As an individual, builder, engineer, or executive, focus on orchestration, workflow redesign, governance, and ROI, not training your own frontier model.
- His core thesis is that the reset is about positioning and distribution more than any one benchmark win, and that we may be about to see the first real loss of OpenAI’s de facto “crown” in AI.
