Why is the investment in startups AI in the growth stage more dangerous and more complex

The bet on the startups of artificial intelligence was not very exciting – or more dangerous. Job occupants such as Openai, Microsoft and Google expand their capabilities quickly to swallow many smaller companies’ offers. At the same time, new startups reach the stage of growth much faster than it was historically.

But determining the “growth stage” in the emerging companies of artificial intelligence is not cut and dry today.

Jill Chase, a partner in Capital, on stage in AI Techcrunch sessions She sees more companies that are only one year old, but they have already reached tens of millions in repeated annual revenues and more than a billion dollars in evaluation. While these companies may be defined as mature due to their evaluation and generation of revenues, they often lack a lot of safety, appointment and the necessary executive infrastructure.

“On the one hand, this is really exciting. It represents this new trend of very rapid growth, which is great,” said Chis. “On the other hand, it is a little frightening because I will pay a billion -dollar evaluation of this company that was not present 12 months ago, and things change very quickly.”

“Who knows who is in a garage somewhere, perhaps in this audience somewhere, a company began to be within 12 months will be much better than this person I invested in that of $ 50 million today,” Chis continued. “So make growth invest a little.”

In order to cut noise, Chis said it is important for investors to feel satisfied with the category and “the founder’s ability to adapt very quickly and a vision about the corners.”

She noted that the AI ​​Coding starting index is a great example of the company “jumped on the state of properly correct use to generate the symbol of artificial intelligence that was available and possible due to technology at that time.”

However, the indicator will need to work to keep the edge.

“By the end of this year, there will be artificial intelligence software.” “In this scenario, what the index possesses today will be slightly less important. The index team must see that future and thinking, well, how do I start building my product even when these models come out and much stronger, the product surface represents these and I can quickly connect this situation and turn to that state of the generation of the code?”