AI Tipping Point?

Recently, many voices have been raised arguing that AI has reached a tipping point.

Most of the arguments are merely trivial:

* “AI requires an entire rebuild of America’s power grid”, which is actually good news for a transition to higher energy density sources like nuclear

* “AI is not cost-effective”; haven’t they heard about economies of scale?

* “GenAi is unreliable”; so was electricity in the late 1800s

* “People overestimated what technology is capable” – do they still remember when the cell phone was launched? In a separate post, I explored Amara’s law: “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run”

* “It is one more investment mania”; if you were in Google’s shoes, where would you invest the zillions of free cash flow? Buy-backs? Would ROI be a concern in an “arms race”’?

* “It is a bubble!”; companies at the start of a tech cycle tend to become overvalued, but it is not anyone’s fault you did not invest in Nvidia (yet)

* “AI most promising feature is to make existence processes more efficient, but estimates of even these efficiency improvements have declined”; have you ever heard of autonomous vehicles, AI-powered robots, genetic mapping, and so on? How can someone be so detached from the tech space?

* “I struggle to believe that technology will achieve cognitive reasoning required to augment or replace human interactions”; that is not the goal of AI; rather it is a powerful technology tool for human civilization enlargement.

Conclusion

We attempt to learn from history.

And the dot-com period offers many valuable lessons.

Thus, while we recognize the revolutionary potential of AI, Fabrica Ventures is not investing in new AI LLMs and applications (at insane multiples) because they are still in the realm of the early Darwinian soup, much like most dot-com companies were.

Instead, we prefer to invest in more established startups which are vigorously incorporating AI features into their offerings.