13 Things to Watch in AI: Macro Trends
We are keeping the momentum rolling with part 3 of our 4-part series on “13 Things to Watch in AI.” In today’s piece, we will dive into a couple of macro trends we are monitoring as the landscape continues to evolve. To catch up, check out our prior posts on technical advancements and product evolution.
07 AI as a GPT (Not the OpenAI Kind)
Navitas believes AI is on the path to becoming a ‘General-purpose technology’ (aka GPT), defined as a technology “that can affect an entire economy… [with] the potential to drastically alter societies through its impact on pre-existing economics and social structures.” Electricity is an archetypal example of a general-purpose technology. Drawing a prescient parallel in 2017, Dr. Andrew Ng claimed, “artificial intelligence is the new electricity.” Like electricity, AI may (1) act as the infrastructure to power other technology applications (as electricity did for lighting), (2) unlock a continuous stream of innovation (as electricity did for refrigeration), and (3) reshape existing industries (as electricity did for manufacturing).
According to analysis by Goldman Sachs, investment in AI could have an even bigger impact than investment in electricity. Examining data on electricity adoption from the early 1900s, investment in equipment and plants peaked in 1918 at 3.5% of GDP, driving comparable growth in labor productivity in 1924 (a 3x+ increase over the prior decade).
Jumping back to the ‘here and now’, in the wake of a record $22B+ in Q4 2023 revenue, Nvidia’s CEO claims AI has hit a ‘tipping point’ with surging global demand across companies, industries, and nations. If Goldman Sachs’ projection of a doubling in AI investment through 2025 proves accurate, we can expect a drastic increase in labor productivity towards the end of the decade, on par with, or even exceeding, the impact of electricity.
08 Where Will Profit Pools Accrue?
As AI becomes increasingly interwoven within the fabric of the economy, who stands to reap the benefits? ChatGPT and other LLMs have captured much of the initial GenAI buzz. But we are still in the early stages of observing how software companies might apply this new foundational technology. Let’s look to the emergence of the internet as an interesting comparison. Computer memory and storage (i.e., cloud services) are the building blocks of the internet age. Starting in the 1990s, the cost of computer memory and storage declined rapidly before starting to decelerate in the 2010s. Although cloud services remain valuable and profitable (see AWS), they have been largely commoditized with minimal differentiation. Meanwhile, software applications running on the cloud (e.g., Salesforce) have captured tremendous value, serving as the portal through which users experience the internet’s remarkable benefits. We may see a similar dynamic evolve as AI matures, where LLMs and other foundational models are increasingly commoditized while Intelligent SaaS businesses are the primary mechanism for delivering value to users.
To keep reading, check out Part 4: Market Shifts.
Over the past 15 years, Navitas has pioneered technological innovation in the built world. With unique expertise spanning venture, technology, and industry, Navitas has earned its market reputation as the partner of choice for start-up founders and leading executives. Artificial intelligence will reveal new opportunities and challenges within the built world. Consistent with the advent of any new technology, there will undoubtedly be winners and losers. As a trusted, engaged advisor, Navitas helps its partners navigate the shifting technology landscape to emerge as category winners, charting the course for the next wave of innovation.
