It’s no surprise that large AI language models are expensive business, but just how costly they are to build hasn’t been clear. As reported by The Information, we now have a glimpse behind the curtain: In OpenAI’s case, the figure totals $540 million in losses last year as the startup developed ChatGPT.
And while revenue is pacing to hit hundreds of millions per annum, expenses are set to grow, too. Specifically, the report mentions that OpenAI CEO Sam Altman may be aiming to raise up to $100 billion in the next few years to develop artificial intelligence so smart it can improve itself. Additionally, OpenAI is keen to acquire fresh data sets for its software, some of which aren’t on the internet at present.
These data sets are a point of contention elsewhere. Companies such as Samsung are wary of their proprietary data slipping into the ChatGPT data pool, as evidenced by a recent ban on generative AI usage with company materials.
Given how massive the cost requirements are for artificial intelligence are (at least in OpenAI’s case), it begs the question whether other AI companies can maintain pace with the trendsetter.
To give deeper insight into the costs behind ChatGPT and OpenAI’s operations, back in 2020, the startup was inclined to spend up to $75 million for Google Cloud services, until Microsoft and its Azure platform became a more attractive prospect for future endeavors.
In The Information’s report, it’s noted that Open AI’s communications and policy chief, Steve Dowling, is planning to depart the company in the near future. Some of OpenAI’s costs have stemmed from hiring professionals from big-name companies such as Apple and Google, so where Dowling’s replacement could come from remains to be seen.
For now, his subordinates are said to be on track to report to Altman directly.