LLMs. Co-pilots. Agents. Two years after ChatGPT took the world by storm, the hype behind artificial Intelligence remains palpable. Major tech companies are pouring billions into sparkling new infrastructure for AI workloads. Utility companies in areas with dense data centers are seeing revenues and power consumption soar. Yet, outside of semiconductor and infrastructure players, few companies have seen significant revenue from AI products. Investors are increasingly left asking: When will these massive capital investments start benefiting software companies? What’s causing the delay?
Where AI is showing up in our world today
First, let’s be clear: AI has been around in many industries for years, including financial services. Quantitative traders long ago embraced techniques that are now classified as "AI" including non-linear regressions and advanced optimization techniques. What excites us now are the rapid advancements in new areas like Generative AI, pattern recognition and deep learning. You’ve likely noticed early use cases of GenAI like Microsoft’s Copilot or Adobe’s AI Assistant. These tools are being integrated into software platforms across industries with new use cases emerging at a rapid clip.
Across other industries, anecdotal datapoints of GenAI value creation are starting to stack up. Walmart recently shared that large language models helped them to “create or improve more than 850 million pieces of data across its product catalog,” completing the project with just 1% of the typical headcount. Amazon’s CEO Andy Jassy shared a similar datapoint when he disclosed that the use of Amazon’s GenAI assistant saved 7,500 “developer-years” of programming time in a recent major software update.
Perhaps the most salient example of AI generating value is hiding under our collective noses in social media, where algorithmic matching on posts like Instagram and TikTok optimizes the content and ads we see. This area has seen tremendous gains in recent years and has helped leading AI adopters in social media improve ad returns and gain market share.
At our firm, we’ve adopted AI with care, led initially by eager adoption of various AI copilots and coding assistants. We have also explored alpha-generating use cases like pattern recognition and sentiment analysis. Looking further out, we are excited about developing custom assistants to superpower our investment and client teams. We believe this is just the tip of the iceberg. Much more is yet to come.
The delay in benefits of AI for software investors
So, if AI is being leveraged across industries, why hasn’t the software sector seen the uptick in growth that investors expected? New revenue streams for software firms haven’t met the hype, leading to stark underperformance for the software industry with a few notable exceptions. We believe this partly reflects Amara’s law: We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run. The rapid adoption of ChatGPT, and the sense of magic many of us experienced when first using it, may have exacerbated elevated expectations.
The reality is that product development and adoption take time. AI development and testing alone can take months, if not years, before a project is ready for deployment. This timeline is further elongated by data governance challenges. As anyone who has worked at a massive conglomerate can likely attest, many large firms have poorly structured and poorly governed data policies, rife with operational silos and inconsistent rules. This complicates development. Compliance and cybersecurity can further stretch timelines. While many firms are seeing real and tangible benefits from AI, getting these projects into production is taking longer than anticipated, creating a mismatch between investor expectations and reality.
While this delay has certainly led to disappointment among software investors, it has not been the only issue. The rapid pace of breakthrough innovation in AI has also sucked the proverbial oxygen out of the room for other technology projects. Some companies have paused major technology initiatives, hesitating to commit to multi\year contracts while still developing their AI strategy. All of this has elongated sales cycles and created a temporary slowdown in software demand.
A window of opportunity for longer-term investors
Surveys have shown that corporations are seeing real, tangible value creation from AI, especially in cost reduction. But AI isn’t just about optimizing existing business models — it’s driving innovation in areas like autonomous cars, healthcare, clean energy, and even entertainment and media. These advancements have the potential to reshape industries and open up new markets. For long-term investors, this makes AI more than a short-term play — it’s a foundation for future growth. Again, like many transformative technologies, AI’s impact may have been overestimated in the short term, but its long-term potential will likely have been underestimated.
We believe the current market anxiety in software creates an opportunity for long-term investors. The project delays are largely temporary in nature and should be resolved as firms update their technological road map and set AI priorities. Recently, we’ve noticed renewed enthusiasm in the software demand environment. As demand normalizes, we expect industry growth to re-accelerate, which should remove the wet blanket from valuations.
Bottom line
A Gartner survey earlier this year found that 90% of CFOs expected increased AI budgets in 2024, with none planning cuts. Yet, it has been clear that many projects have faced delays due to complexity, data governance issues, security concerns, and a scarcity of talent. This has created an air pocket in an industry that is usually known more for "eating the world" than for booms and busts. Just as the internet didn’t take off overnight, AI still has some development before it starts to deliver significant results for investors. We believe those returns are coming, and the impact is on the horizon if investors can remain patient. Now is the time for investors to position themselves for the shift from hype to impact. As AI projects move from development to deployment, those who stay the course or increase their exposure to the right opportunities are likely to see meaningful returns ahead.
Dave Smith is head of technology investing at Bailard. He is based in San Francisco. This content represents the views of the author. It was submitted and edited under Pensions & Investments guidelines but is not a product of P&I’s editorial team.