The private market valuation of OpenAI indicates that the runway for Artificial Intelligence may be much larger than street analysts cluster around.
In the most recent round of private capital raising, OpenAI has achieved a valuation in excess of $150 billion. The proforma revenue forecast for the maker of ChatGPT, now that it is converting from a not for profit to a profit enterprise, is set to advance from $4 billion in 2024 to $11.6 billion in 2025. That virtual tripling in revenue sounds mindboggling, but as improbable as that seems to investors, it pales in comparison to 5 year targets. OpenAI intends to build on that revenue and target a $100 billion sales forecast in 2029, a compounded annual growth rate of about 90%.
Revenue forecasts with such staggering growth rates imply that the AI industry is already advancing from open competition to monopoly status.
Clearly, OpenAI will be conveyed with such status and until such time as the company goes public, it will be exclusively the investment vehicle for just a handful of initial backers, including Microsoft. That leaves individual investors and smaller institutional accounts scrambling, scratching their heads, to obtain a public proxy vehicle/s for participation.
A high margin, monopolistic/oligopolistic sector with high growth rates historically indicate that small or mid-cap investment opportunities are highly unlikely to be found.
The barriers to entry in AI are humungous. The capital equipment required is super expensive. Gathering extensive and usable databases is super expensive. Name recognition is essential. Security of data is paramount.
“Small cap AI” and “Mid-cap AI” are oxymoronic statements. This field will be the largely, exclusive, domain of a handful of ultra large cap companies and nobody else. I surmise that a company requires a market cap of at least $100 billion just to be permitted a peek at the sector.
Investors, both large and small, should be wary, as the prospect for getting burned on non-participants, or through misallocation of scarce capital into marginal players that will produce less than optimal returns, based upon marketed pretense as being a part of the AI boom, is far, far, greater than the number of meaningful business winners suggest.
There will be promoters declaring AI to be the new wild west, with opportunity for all.
Not so.
Laws have already been established, marshals hired, judges at the ready, territories carved up and boundaries surveyed. These AI landowners want ALL of the business and have paid their admission fees. Remaining scraps will only be useful as compost.
The market is already carved up, now largely finite, but most have yet to connect the dots.
Just as with smartphones, where there is the Apple platform and the android platform, there will likely be just two, maybe three legitimate AI creators, platforms, of any consequence. Now that OpenAI has opted to convert from a cooperative to a for-profit entity, AI won’t be free, nor will it be cheap. Corporate AI will be costly and when sold as a service, will produce reliable, super high margin, revenue streams for the AI producer(s).
Next, there will be the datacenters storing the AI and Microsoft has already indicated their capital design upon being #1 in that category. There will be some pretenders in play, but AI is built upon the acquisition and storage of data for algorithm training. The company that has the most dedicated servers and the greatest amount of stored data will have, all else being equal, a competitive advantage. Microsoft is early/on trend in this field. There isn’t a business opportunity for those that are late, because AI improves with increased data. Who will to pay to acquire old-gen AI from a smaller vendor?
There will be the manufacturer of the AI chips, and right now there is the brand name company (Nvidia) as well as the actual producer of the chip itself (Taiwan Semiconductor).
There will be server demand, as data needs to be housed.
Finally, there will be a couple of companies training buyers on how to best utilize the “AI As A Service” (AIAAS), because most businesses haven’t a clue what do actually do with artificial intelligence to better their corporation. Everyone is running towards it, under peer pressure, and as yet, almost nobody knows why. The winner in this service subsector will be the company deemed to be the most secure, that trains users but does not simultaneously poach corporate data for resale and that has a large, high quality install base to convey credibility.
As for every business putting out vacuous statements of intent on the adoption of AI, they will be the actual consumers, not the producers.
A company that purchases AI from OpenAI, then hires Microsoft to help store that data and that uses Palantir to teach them how to use the artificial intelligence for their needs; that company isn’t an AI business at all, they are just the consumer.
A bank, payment processor or finance company announcement of the creation of an AI department to roll out the model doesn’t make any of these firms an AI play; they are just reporting that they are customers, adding another line to the expense column of their fiscal reports.
A retailer rolling out AI algorithms to improve the cart shopping experience online isn’t an AI company, they are a consumer.
A utility selling electricity to an AI data center isn’t an AI company, they are a regulated supplier of energy, that’s it.
A call center firm using AI bots isn’t an AI company, they are just a consumer of AI.
An engineering company using AI to supplement their personnel isn’t an AI company.
The list of who IS NOT an AI company, nor will ever be, is incredibly long.
The list of who IS an AI business is incredibly small.
That is why OpenAI has produced a 5 year forecast guiding towards a 25X increase in revenue growth. There isn’t a misprint, nor is this a typo, I didn’t incorrectly state a 25% revenue growth rate. No, OpenAi has declared that business will profoundly accelerate, a massive trend, and the trend may lead to a 25X revenue increase by 2030.
If OpenAI has modeled such a growth rate in revenue, analysts, on the whole, might be woefully wrong on forecasts for related, ancillary businesses in this business chain.
At some point, relatively soon, OpenAi will likely go public, potentially with a valuation in the $250 billion-$300 billion range, maybe higher. Such a valuation shock for investors may produce a material uplift for any actual business, legitimately on trend, in their participation within the AI business model.
High value targets have been identified within the OpenAI forecast. Intrepid investors need to to connect the dots for themselves, the analytical street is behind trend.
Leave a Reply
You must be logged in to post a comment.