The Race For AI Partnerships Is On And It's Going To Be Hilarious
For us. Not so much for them.
The next 12 to 24 months are going to be wild when it comes to investing in AI-first businesses or startups. Investors don’t know what they should pay attention to, and hype-driven FOMO is stirring the pot. The combination has led to unbelievable valuations in the past and probably will again now.
Why does this matter? Businesses will be joining the ranks of VCs this time around. A few forces are combining to create a strange cycle. I will start with why companies are jumping in.
Businesses outside tech don’t have the internal capabilities to develop large models. Many will explore building a large-scale AI center of excellence but quickly abandon that idea. The main reasons:
It’s expensive, and traditional businesses aren’t structured to scale returns faster than the costs. The monetization paradigm is novel, and business models must transform to support it.
Models aren’t built with digital development cycles and the stability that can come with them. As a result, they don’t fit on a traditional product roadmap.
There’s not enough talent available. The talent grab is on for applied researchers, technical strategists, people who can convert research artifacts into products, and people who can monetize those products.
Most businesses will develop data science capabilities, but not to the level they would need to build something like ChatGPT. Companies will be forced to partner or invest in businesses that have.
How Partnerships Played Out In Other Industries
Automotive companies underwent the same cycle to bring autonomous vehicles to the markets. Mercedes partnered with NVIDIA in 2018 to develop the architecture for next-generation intelligent vehicles. Mercedes is the first to bring level 3 autonomous vehicles to market in the US. Hyundai has also announced plans to bring level 3 vehicles to the US this year.
Companies like Ford partnered with startups, and their track record has been mixed. They recently ended their partnership with Argo AI and announced plans to step back from moving further into developing fully autonomous vehicles. Their initial investment of $1 billion was made in 2017, and they put another $1.7 billion into the program over the next 5 years.
Competitors will drive companies to move into increasingly complex AI product spaces. Just as in the automotive space, assessing partnerships will be a critical success factor. Companies cannot move forward alone.
Even Microsoft and Google pursued partnerships to accelerate their AI product journeys. Microsoft’s success is due, in large part, to OpenAI. They chose a winner back in 2019, and that has paid off this month. The $10 billion investment has already boosted its market cap by over $200 billion in one month.
This year, opportunities at scale and threats from rivals will lure companies into partnerships. Few will have the necessary tools to evaluate those partners, so valuations will be all over the place. We will look back in 2-3 years and see some tragic decisions based on flawed metrics.
History Repeats Itself
The same pattern is playing out for AI startups that hit social media apps. As they went public and investor scrutiny increased, CEOs and founders had to teach investors an entirely new set of metrics. We’re about to go through that same churn and uncertainty again.
The one thing that became obvious during the last cycle is investors are confident that they can find ways to quantify a brand new business model using existing metrics. This time around, investors will try to stuff AI products into a digital software or platform paradigm. It won’t work because the monetization is entirely different. Evaluating the business based on subscribers or ARPU won’t be forward-looking.