Nagrani highlights the fact that during the hype cycle phase of a new technology, caution often goes out the window. Venture capitalists (VCs) feel the pressure to deploy their raised capital and fear missing out on the next big thing. This same dynamic plays out within companies, with leadership greenlighting investments in the name of AI that may be nothing more than misplaced hope.
While it’s true that large language models (LLMs) and AI can be game-changing technologies, Nagrani advises companies to approach their investments with a clear-eyed perspective. He offers three essential tips for preventing wasted spend on AI:
1. Understand total cost over time: Before saying yes to an AI project, companies should consider the cost of the needed resources both currently and over time. It’s important to look beyond just the data science team and account for engineering, DevOps, QA, product, and SysOps time. Failure to invest in ongoing support can leave projects abandoned and unfulfilled.
2. Ask why someone else can’t do it: Companies should question why they are uniquely positioned to tackle an AI initiative. If the project is likely to commoditize over time, without a scale or cost advantage, it may not be worth pursuing. Instead, focus on areas where there is a defensive moat, such as preferential access to data or proprietary insights.
3. Make a few bets you’re willing to follow through: The simplest and easiest bets are those that improve the products a company is already making. These bets can provide momentum and scalability. Companies can also consider moving up or down the value chain or expanding laterally into other sectors. The most challenging but important bets may require cannibalizing the current business with new technology, but failing to do so could mean losing out to competitors.
By following these tips, companies can ensure that their investments in AI have the best chance of success and avoid wasting billions of dollars. While the hype around AI is real, it’s important to approach it with a rational mindset and carefully consider where to place bets.
In conclusion, AI has the potential to revolutionize industries, but it also comes with risks and potential wasteful spending. Companies must navigate the hype cycle and make strategic decisions when investing in AI. By understanding the total cost over time, questioning uniqueness, and making calculated bets, companies can maximize their chances of success in the AI landscape.
