Silicon Valley has been the epicenter of innovation and technological advances for decades now. With the progress made using generative AI in the last year, this isn’t likely to change any time soon. Large and small companies as well as many startups are hopping on board the AI train hoping to make it big. But things aren’t always as easy or as predictable as they seem. Certainly, it’s evident that generative AI will change the world as we know it. However, the AI hype bubble may be exaggerating when it comes to opportunities to make a profit. Over the last year, AI startup investments have been pouring in, but that pipeline may soon stop. As AI startups begin to call it quits, investors are likely to take a much closer look at a company’s potential.
(What kind of AI do you want in your future? Read this Bold story and decide.)
The problem with creating, developing, and maintaining generative AI systems involves cost more than anything. The amount of AI startup investments required is typically in the billions, not millions. This means the AI investment game has a great deal more to risk than previous tech booms in Silicon Valley. Likewise, few companies have figured out how to generate the amount of revenues required to provide a return on investment. While tech giants can tolerate such ambiguity and massive AI investments for a while, startups can’t. They simply don’t have the luxury of time or deep pockets. Thus far, they have been riding the wave of the AI hype bubble. It seems now, though, that at best a slow leak in that bubble may be occurring.
Big Money, Few Returns
In assessing the size of AI startup investments required, industry totals offer a great snapshot. According to some reports, investors sank a whopping $330 billion into 26,000 different AI startups in the last three years. By comparison, the amount invested in the three-year period prior was about a third of this. This degree of funding activity supports the growing AI hype bubble in recent times. But the returns on these investments have been minimal. In considering on AI startups and not tech giants, only a few have actually generated revenues. And when comparing these revenues with investments, they reflect a tiny fraction. Therefore, sustainability is in question if such investment amounts continue to be required.
Investors are now starting to examine business viability for these AI startups given that some are closing their doors. Most have assumed that there is some type of long-term model where generative AI systems can generate large sums of money. Recent discussions of smaller generative AI platforms that tailor to specific tasks offers one such model. Generative AI assistants for personal and/or professional use are other possible options. But realizing these longer-term gains may pose to be difficult, especially with large AI startup investments are needed. This is why the AI hype bubble is contracting to some extent lately. And its why access to additional funding may prove difficult for AI startups going forward.
Recent AI Startup Struggles
Amidst these investor concerns, several AI startups are known to be struggling. Recently, Inflection AI folded, recognizing its inability to sustain operations long term. As far as AI investments go, the company raised roughly $1.5 billion. But at this point in time, it has failed to generate any significant revenues, and its ability to do so soon is unlikely. Anthropic, another AI startup, has announced similar struggles as well. They are currently experiencing a gap of $1.8 billion between costs and sales. They have raised about 200 million in sales revenues but is spending $2 billion annually. Then there is Stability AI with several resignations recently from their company including top researchers. They too are upside down with $96 million in expenses but only $60 million in income revenues. Given these developments, many investors have become disillusioned in the AI hype bubble at least as it pertains to startups.
(One of the biggest problems with AI is monetizing its use–read all about it in this Bold story.)
Of course, revenues from generative AI thus far aren’t that great even for more established AI companies. Open AI only earned $1.6 billion in 2023, which is significant compared to smaller AI startup companies. But it’s also likely that Open AI’s operational expenses are tremendously higher as well. Though Open AI’s expenditures are not known, it’s believed that these far exceed revenues. This is particularly true given the legal challenges the company is facing over copyright concerns. Open AI is in a large way responsible for the growing AI hype bubble. But if its costs were known, this hype might not be quite as grand as it has been.
The Bigger Challenge for AI Startups
While coming up with a viable revenue model will be essential, AT startups have bigger hurdles to conquer. Most realize that competing with the tech giants in the field, massive AI startup investments will be required. These investments were more readily available as the AI hype bubble was expanding. But as realism sets in, these funds won’t be available in needed amounts much longer. This is why some of the most competitive companies are aligning themselves with the tech giants for security. It’s well known that Microsoft fueled Open AI’s growth with a $13 billion investment. Likewise, Anthropic has now accepted support from both Amazon and Google to the tune of $7 billion. This is the amount of AI startup investments needed to compete moving forward. And many AI startups simply won’t have this capacity.
Unfortunately, advancing generative AI is a costly endeavor. For one, the semiconductor chips needed are in limited supply and are quite expensive. At the same time, AI training demands a tremendous amount of data, which also can be pricey. Not only does this involve access and time, but ongoing battles over AI training and copyright laws pose threats too. And then there are additional risks of AI systems making errors, which could undermine revenue generation. Thus, the need for large AI startup investments isn’t going to go anywhere anytime soon. Hopefully, the AI hype bubble won’t burst, and such funding will persist until startups can turn the corner. This remains to be seen, however, as investors begin to get increasingly cautious.
It’s perilous to program AI with biases–read why in this Bold story.
