Welcome back! After a brief hiatus for eye surgery (nothing too serious, but enough to warrant a pause), I’m here to talk more about the ever-evolving world of artificial intelligence.

Lately, the buzz in generative artificial intelligence (GAI) seems to have firmly hit the off ramp from pure innovation into incremental iteration. Sure, we’ve still seen the occasional technological strides — although they’re often more about hype than real heat. But the real story is the evolving discussion around innovation in large language models (LLMs). It’s becoming yet another tech-industry melodrama, with the would-be kings all eyeing OpenAI’s throne. And when all your rivals constantly measure their success against yours, it’s a clear nod to who is currently dominating the market.

Since the craziness of the Sam Altman firing and rehiring, the last few weeks of drama has been mostly (but not completely) focused on Google’s latest entrant, Gemini. Their new model has stirred quite the conversation, presenting itself as the new challenger on the block — ready to upset the status quo. However, more than a pinch of skepticism has been warranted, especially after Google’s recent demo ended up with more than a little bit of good old fashioned stagecraft. It underscores the intense desire among tech giants to claim dominance in a market heavily influenced by OpenAI/ChatGPT’s innovation. For Google, a pioneer in the Transformer technology that made generative AI possible, this must feel more than a little bit galling.

The challenge for Google, and indeed for any sufficiently large corporation looking to innovate, is akin to an old dog trying to teach itself a new trick. Their attempt to interlace AI across their product range, though ambitious, doesn’t seem to be flowing as naturally as Microsoft’s collaboration with OpenAI (although they’re happy to charge the same). Despite being the more mature company, Microsoft seems to be taking Zfresher perspective. Meanwhile Google seems to be getting hamstrung by its own suite of legacy products and interfaces.

What we’re witnessing isn’t just a race for technological superiority; it’s a battle for innovation within the confines of corporate legacy. And if they do reach the top, what’s next? Once Google holds the lead technologically, are they even capable of pivoting to genuine innovation after playing catch-up? The nature of software development means we’re fast approaching an asymptotic curve where even the biggest tech investments only yield modest gains. And even if they can reach the summit, let’s not forget the hard economic realities. Everyone is at least partially subsidizing the real costs of AI computing. Google might be feeling generous now, but once they’re leading the pack, expect a quick shift towards profit maximization on the back of the consumer.

Thankfully, the AI market isn’t going to be dominated by a single climber for quite a while. With players like Meta, Amazon, and a slew of open-source models in the fray, we’re shielded, at least for the time being, from a monopolistic AI landscape.

Different LLMs may gain a temporary edge, but beneath this facade of “pure power” lies a complex mix of marketing, strategy, and perhaps a bit of showmanship. It’s not just about who has the flashiest AI; factors like integration, customization, and user experience will play pivotal roles.

As we step into 2024, we’re transitioning from the breathless innovation of the previous year and entering into an era of incremental progress as we ascend up the asymptotic curve. The broader market might not fully appreciate these nuances, but the landscape is undoubtedly getting steeper as we climb to ever more dizzying heights. So, keep your wits about you, don’t get swayed by the superficial, and remember – in the AI world, even the mightiest can lose their footing if they start to believe their own hype.