AI investors are navigating ambiguity

? AI investors are treading murky waters with an uncertain future ahead. Stakeholders are elated over the prospects of AI, with many seemingly underestimating the unpredictability of the technology. Concerns have been piling up about an AI bubble that could burst. With so much money flowing into startups and companies, will AI meet financial expectations?

Investors seem to be charging towards a profitable AI future faster than the technology itself can keep up, with expectations that large language models will reach full potential and develop the long-awaited superintelligence — or Artificial General Intelligence (AGI). Just in the first half of 2025, 50% of venture capital was spent on AI startups, according to Business Insider.

Investors are “overexcited” about AI, with many risking enormous loss, OpenAI CEO Sam Altman predicts. At the time of making these claims, OpenAI had just valued itself at more than USD 500 bn, with a USD 6 bn deal from investors like Softbank and Thrive Capital, according to CNBC. With the exception of big companies like OpenAI, it seems AI is not making huge returns on investment as promised, with only 5% of companies that embraced agentic AI seeing returns, a recent MIT study showed.

Why is the AI world so unclear right now? According to General Catalyst CEO HemantTaneja, the unpredictability stems from the models themselves. As the models keep getting smarter, the applications and infrastructures that they’re built on are rendered obsolete. It’s hard to discern which models are sustainable — and in turn, where to invest. Taneja references the now-obsolete companies that were funded when GPT-3 launched because GPT-4 superseded its functionalities.

Current valuations are fitting within the large landscape of AI. Investors in AI like Meta, Google, Amazon, and Microsoft are not taking huge risks when pouring their money into startups or AI data centers. AI companies now boast considerable earnings and strong cashflows that can sustain losses, according to ARS Technica. This is why the AI bubble could only witness a gradual deflation as long-term impact begins to manifest, rather than a sudden pop.

Taneja, however, also remains optimistic, predicting that much of AI development will spring from the service business companies he’s currently investing in, and their own transformations with AI. He also predicts the Mergers and Acquisitions market will be an interesting one for AI companies.