Intel is taking some heat for not investing in OpenAI in 2017. It's easy to look back on that from 2024 and prompt ChatGPT to write a scathing takedown of Intel's lack of vision. Instead, I looked into the details.
The reported terms of the deal were 15% stake for $1 billion in cash, with an additional 15% stake in OpenAI if it made hardware for the startup at cost price. That is, significant portions of their facilities and personnel would be generating zero profit.
Seven years later, OpenAI is still not profitable, losing gobs of money every year and committing to spending even more over the next several years.
Given that Intel was concerned with making their money back in the near future and OpenAI was nowhere near that (and still isn’t), this seems to have been a reasonable decision.
The strategy to focus on their core business looked incredible at first: in the years immediately after turning this investment down, Intel achieved near-all-time high stock price and all-time high EPS and revenue. Not bad. This strategy might have even enabled them to outperform during the supply chain mess of 2020 & following.*
And yet, the last couple years have been rough (to say the least) for Intel: “The thing is, if you don’t invest in the future, or see fundamental changes in the market coming, then you don’t have to spend as much; if you don’t have to spend as much then you can increase margins.” And remember: had they chosen to produce massive amounts of profit-less hardware, it's unlikely* they would have been able to generate the cash necessary for survival.
The miss wasn’t just a lack of investment in or partnership with OpenAI, it was much deeper than that. They avoided shifting the business model to match the emerging landscape, not just the investment thesis.
A few trailing thoughts after all this:
It's easy to grab a headline, knock someone down, & miss the real story
You evaluate decisions based not on the outcome, but on the process you went through to make the decision using the information you had at the time. Post-hoc evaluations are unfair, arrogant, and lead to worse decision-making in the future.
ChatGPT (as seen below) doesn't even give a full & accurate assessment of what happened. Do your own work.
* Hypotheticals aren't really fair, of course, which is kind of my point: it's easy to say, "If only Intel had invested!" but we don't know where Intel — or OpenAI for that matter — would be right now if they had done so.
•••
For what it’s worth, I asked ChatGPT: “what can you tell me about intel's decision to not invest in openai in 2017?“
In 2017, Intel decided not to invest in OpenAI during one of its early funding rounds, a decision that has since been seen as a significant missed opportunity. At the time, Intel was heavily focused on its core business areas, such as manufacturing processors and other hardware. The company may not have fully grasped the potential impact of AI and large language models or may have been hesitant to invest in a field that was still emerging and relatively uncertain.
OpenAI was seeking strategic investors who could provide not just capital but also deep expertise in AI, computing power, and a commitment to advancing AI technology. Intel's decision might have been influenced by various factors, including its assessment of the financial and technological risks, as well as its overall strategy and focus at the time.
In hindsight, Intel's decision not to invest in OpenAI has been contrasted with the moves made by other companies and investors who did back the AI startup. OpenAI has since become one of the most influential organizations in AI research, with significant advancements in machine learning, natural language processing, and other AI technologies.
This decision is often discussed in the context of how companies evaluate emerging technologies and the long-term implications of those decisions. For Intel, which has faced challenges in diversifying its business beyond hardware, missing out on an early stake in OpenAI represents a noteworthy moment in its corporate history.
My follow-up: “had intel invested, what kind of return would they have gotten by now?
If Intel had invested in OpenAI back in 2017, the potential return on investment (ROI) would likely have been substantial, given OpenAI's remarkable growth and the increasing value of AI technology. However, the exact return would depend on several factors, including the size of the investment, the specific terms of the investment deal, and how much equity they would have acquired.
Here are some factors to consider:
Valuation Growth: OpenAI's valuation has surged over the years, especially with the success of models like GPT-3 and GPT-4. By 2023, OpenAI had been valued at around $29 billion, following a significant investment from Microsoft. If Intel had invested in OpenAI at a much lower valuation in 2017, the increase in valuation would have yielded a very high return.
Equity Stake: The actual ROI would depend on the equity stake Intel would have received in exchange for its investment. For example, if Intel had invested $100 million in 2017 at a $1 billion valuation, they would have owned 10% of the company. With the company later being valued at $29 billion, that 10% stake could potentially be worth $2.9 billion, resulting in a 29x return.
Strategic Benefits: Beyond the financial return, Intel might have gained strategic benefits from being closely involved with one of the leading AI research organizations. This could have influenced Intel's AI strategy, provided early access to cutting-edge AI technologies, and strengthened its position in the AI market.
Potential Dilution: Over time, additional funding rounds would likely dilute Intel's stake unless they participated in follow-on investments. This dilution would reduce the percentage of ownership, although the overall value of the stake could still grow due to the increase in the company's valuation.
Given these factors, it's clear that Intel's return could have been quite significant, both in terms of financial gains and strategic advantages, had they invested in OpenAI during its early stages.
