Halicin, a drug identified by AI and discovered by researchers at the MIT Jameel Clinic that successfully kills antibiotic-resistant bacteria, exemplifies the potential of AI to achieve unexpected results. Such technologies not only open up new possibilities for human health, but as a Forbes article notes, come at a high price. So how can, and should, AI be funded?
Excerpt
Clearly, AI can produce unexpected results. Take just two popular examples:
• DeepMind’s AlphaZero taught itself to play chess with unorthodox tactics. It sacrificed pieces that humans consider vital, including its queen. Its self-taught domination led grandmaster Garry Kasparov to declare, "Chess has been shaken to its root by AlphaZero."
• MIT's halicin project triumphantly killed strains of previously resistant bacteria. Traditional R&D would've been prohibitively expensive. Instead, AI learned and identified attributes that had eluded human conception.
When such unimaginable results are possible, it’s no wonder AI budgets are up. But it begs the question: Should the cost be borne 100% by IT budgets, or should AI funding come from unconventional sources?
My colleagues and I are involved daily with organizations that are already using or are considering generative AI (GenAI) systems like ChatGPT and Microsoft’s Copilot. We typically interact with CIOs and their IT staff.
Broadly, they’re divisible into two camps:
1. Those ready to invest in the tech without deep ROI evaluation. We’ll call those the early adopters.
2. Those curious about the tech but want to evaluate ROI first. Let’s call them the risk-averse.
Whether risk-averse or an early adopter, a CIO is right to scrutinize the investment. Enterprise-grade GenAI can cost up to $30 a month, per user. That’s a significant increase over existing productivity app costs and is challenging for squeezed IT budgets. After all, few had planned for an additional line item like that just 12 months ago.
Organizations that evaluate too long will miss out on the early mover advantage. Those that aren’t yet evaluating at all may indeed be left behind by their nimbler competitors. Organizations that spread AI’s costs across the lines of business (LOB) that will leverage it can seize a competitive advantage. The remainder of this article provides evidence of how and why.