AI Ecosystem – Starter’s Orders

The race has just begun

• The contenders are already jostling for position when it comes to ownership of the emerging AI Ecosystem where OpenAI is in the lead, but its internal foibles have given everyone else and especially Meta a leg up.
• The smartphone ecosystem is already well established where Apple, Google and Tencent are the clear winners, but the AI ecosystem is barely getting started meaning that the space is wide open.
• The smartphone ecosystem is worth trillions of dollars in terms of value creation and the AI ecosystem could easily be as valuable over time even after the current bubble bursts.
• This is why there is great interest from start-ups, giant tech companies and investors to carve out a piece of the opportunity.
• Any digital ecosystem is all about the network effect where delighting users, making it easy to use and attracting developers is the key to success.
• OpenAI made an early move into this space (see here) with a platform for generating custom AIs based on GPT and a store where developers can make their creations available for sale.
• However, OpenAI self-immolation (see here) has clearly set its plans back somewhat as the GPT Store launch has now been delayed to early 2024 and the company has done an untold amount of damage to its reputation.
• This is critical because if developers view OpenAI as flaky and unreliable, they will be less inclined to base their businesses on the GPT platform giving an opening to the increasing number of rivals snapping at its heels.
• Its rivals are jumping at the chance and Meta in particular is very keen not to miss out the second time around.
• Meta’s strategy has been to harness the open-source community by making its LlaMa foundation models and their weights available to anyone who wants them for no charge.
• This would seem to be a foolish thing to do but the return for Meta is that its foundation models are now the standard for open source development.
• This also means that anything that is developed in the open source community can be used by Meta in its products which means that Meta has outsourced a lot of R&D for the cost of the foundation model.
• This has caused a lot of critics to call the open-sourcing of the foundation model dangerous but at the same time, they also say that Meta’s foundation model is second-rate.
• Consequently, I don’t put a lot of weight on these criticisms as they look like competitive marketing to me given their logical inconsistency.
• The emerging AI ecosystem is currently made up of new players in the ecosystem game as neither Apple nor Google have yet made a serious move in this direction.
• I remain confident that Google has the skill set to do this and will move shortly, but I have much less confidence when it comes to Apple.
• Siri has been second-rate for years and Apple’s secretive nature has been a hindrance in what has been an open and collaborative environment.
• However, as money starts to be involved, everyone is becoming more guarded in terms of what IP they share and what happens to their algorithms and so one could argue that the market is moving towards the way that Apple does things.
• I am not sure that this is going to be enough to allow Apple to catch up and so I think that Apple remains one of the weaker contenders at the moment.
• The other backmarker at the moment is Amazon whose AI has also been poor for many years, but it has signalled its intent to deal with this through its investment and possible acquisition of Anthropic.
• The race has barely begun but already there are thrills and spills as OpenAI has squandered its early lead and Meta has closed some of the gap and doubles down on its open source strategy.
• As it was in the smartphone ecosystem, the early leaders were not always the winners and there remains everything to play for.
• Nvidia remains in a very strong position when it comes to AI training but if the control point for AI shifts away from silicon chips then its grip may weaken as all of the current AI ecosystem contenders are keen to diversify their hardware requirements.
• Nvidia remains priced for perfection but while it continues to blow expectations out of the water, its valuation is likely to hold.
• I continue to think that there are much better places to look in the technology sector.