Tech companies are investing billions in AI but not seeing proper ROI. Microsoft’s $19B concerns analysts, potentially making AI appear pretty useless so far, while companies are using the “it’s a long-term investment” argument.
As we all know, major tech companies have jumped on the AI gravy train, so to speak, with their billions of dollars. However, compared to other forays these companies have made, such as new products, cloud technologies, or other applications, AI is turning out to be quite the less profitable bid.
These big investments concern analysts and possibly also stakeholders. For example, Microsoft spent $19 billion on AI in just one quarter. The return on investment is nowhere to be seen (not just for Microsoft, but for other companies as well).
The story was published on Futurism, where it’s also said that Google would have spent $49 billion on AI by the end of 2024.
Generally, AI tech is very expensive to develop and requires specialized chips and data centers (which has ballooned Nvidia’s market capitalization to new heights). What many believe is that AI is nowhere near useful — yet.
On the other hand, these tech companies are justifying the investments by saying that it’s a long-term game. That is true to some extent, but the technology of this kind when money is being poured into it left, right, and center typically takes off by now. And by taking off I mean starts generating valuable revenue for the companies to keep justifying new rounds of investments.
That doesn’t seem to be happening right now. Generative AI tools are simply not generating enough even with all the APIs and premium subscriptions.
To me, it feels like the tech itself is its own biggest enemy.
On one side, you have casual internet users. They are okay with the free versions. After all, the free ChatGPT or Gemini can do the majority of tasks that generative AI is so hyped up about, like holding a conversation or diagnosing cancer for you (let’s face it, we all kind of have it according to the internet).
On the other hand, more advanced users can utilize the many free, open-source models and set up their own AI chatbots or tools on their machines natively. Open-source AI has grown by leaps and bounds, and that’s clearly an interesting case study in the larger story.
Who, then, buys the subscriptions and APIs? Just people who need specialized tools where it’s okay to shell out this amount because the tool itself will be making money? These wrappers, tools, and other kinds of AI-integrated programs are not many. At least there’s a lack of really useful ones, but we do see new ones cropping up almost every day. These include everything from AI writing and detection websites to paid AI voice generation or graphic novel creators.
That kind of only leaves people who like to be on the cutting edge all the time without having the technical interest or bandwidth to set up their own LLMs. And you cannot make AI tools a net profit maker just by relying on this segment.
