Gensler’s AI Washing

When the SEC announced actions against Delphia and Global Predictions on March 18th for their claims of using artificial intelligence (AI), it initiated a broader conversation: what really qualifies as AI? This isn't just regulatory nitpicking; it strikes at the heart of how we define and understand AI in the financial industry and beyond.

AI is a term that's tossed around with the abandon of confetti at a parade, often used more for its marketing sparkle than its technical precision. The saying goes that "machine learning is done in Python, and AI is done in PowerPoint," suggesting that what is often touted as AI is merely dressed-up math and statistics.

Through years in the trenches of applied math and statistics, I've come to see AI as a duality: the analytical and the generative. The analytical is the workhorse, applying rules such as "if the temperature is above 20 degrees, wear shorts." It's logical, it's reliable, but it's hardly what one might call revolutionary.

On the other side is generative AI—the creative maverick. This AI doesn't just follow rules; it writes its own symphony. It's the kind capable of crafting new content, whether that's in writing, art, or even music, and this is where things get interesting.

The real juice, I believe, is where AI can bring a human touch to the table, generating novel insights and strategies. This is not about replacing jobs but enriching them, adding roles like prompt engineers and ethics monitors as analyzed in the FT on March 26th (Financial services counting on AI for a productivity boost). At Eqitron, we're actively exploring these possibilities, looking at how generative AI can bring fresh ideas to the fore, potentially even reshaping the role of a researchers, analysts and portfolio managers.

In essence, while the term AI can be a smoke and mirrors game, the SEC's recent actions remind us to look behind the curtain. The future of AI, particularly in sectors like finance, may well hinge on our ability to harness the generative side of the spectrum, moving from "what if" to "what could be".