Generative AI seemingly took over the tech world consciousness following the release of ChatGPT in November 2022, dominating both news cycles and many companies’ earnings calls last year. For nearly all of 2023, the biggest news stories revolved around OpenAI, and the biggest question on every company’s mind was how it would capitalize on it. The options were to innovate with AI or risk being left behind the promise of AI: more efficiency, more effectiveness, and lower costs.
Now that many companies have experience building, integrating with, or using one genAI tool or another, it’s time for a closer examination of AI and its applications. We saw this same cycle with the introduction of the internet decades ago and the Dot-Com Bubble in the late 1990s. While there is no doubt that AI will continue to evolve and impact nearly every human activity, this is the year for companies to take a closer look at how they’re incorporating AI to drive real impact.
What happened in the 90s
In the 1990s, electronic communication was very basic and very limited. It wasn’t until the invention of the World Wide Web that the Internet became more accessible to the general public, being used for information sharing outside of strictly academic, government, and research institutions. As internet usage boomed throughout the 1990s, a multitude of internet startups emerged, covering various sectors such as e-commerce, online content, and app development.
Investors were drawn to the potential of the Internet and its emerging technologies. Many companies with little or no profits saw significant increases in their stock prices and the valuations of internet companies became disconnected from their underlying fundamentals. Traditional metrics like price-to-earnings ratios were often ignored, and investors were more focused on potential future earnings and growth.
Things took a turn in 2000 when investors began to realize that many internet companies were overvalued and faced significant challenges in turning a profit. This led to a rapid sell-off of technology stocks, resulting in substantial losses for investors and the notorious Dot-Com Bubble.
While most of these early internet companies failed, others soared. Microsoft, which was founded before the internet, was a major beneficiary of the Internet Era and is a big proponent of generative AI today. Amazon, Google, Cisco: the survivors are some of the world’s biggest companies today. The rest are lost to time.
However, as renowned futurist Roy Amara once said, “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” With this in mind, we should be in this AI race for the long run – but at what pace?
A look at AI in 2024
Tech is a sink-or-swim world. That’s why this year is all the more important for companies to slow down and consider the real implications of generative AI. Jumping on the bandwagon only to create another bubble will hurt more when it bursts. The allure of genAI is in its potential, and potential is the thing it has in excess – but it’s unlikely that 2024 will bring any huge breakthroughs. Rather, this is the year companies should spend giving it a closer examination to figure out exactly what its place is in their industry.
The next logical step is fine-tuning AI for specific applications and addressing industry-specific needs, along with more regulatory guidelines, which have been a long time in the making. The EU and China are leading the pack in regulating AI, working to establish ethical guidelines for its development and deployment. It could be indicative of more regulations to pass in other countries as well, including the US and UK.
This means that companies will have to adapt; but as AI becomes more specialized, we’ll also begin to see more seamless integrations into everyday applications as we inch closer to realizing its full potential.
Dependencies on the AI journey
AI development relies on other industries, like public cloud and colocation providers, to power the equipment that allows them to train AI. 2023 was a record year for data center leasing, but because it takes 24 to 36 months for new data center builds to come online, the majority of data center leasing for AI workloads still needs wait for public cloud and colo providers to transform their backlog into operational infrastructure.
While cloud and colo will continue to grow with both natural enterprise and hyperscale growth and increasing AI demand, providers of those services are optimizing their infrastructure to accommodate. The tech industry is incredibly dynamic and responsive to new trends, and the AI journey, likely, will not negatively impact other industries.
Despite the downturn and the collapse of many companies during the Dot-Com Bubble, the technology sector recovered and evolved. The lessons learned during this period have influenced subsequent investment strategies, with a greater emphasis on sustainable business models and profitability and can be reflected in the state of AI development today.
