Since they landed in everyday life, generative Artificial Intelligences such as ChatGPT have had the air of a computing revolution: conversations, advanced functions, complex responses, text generation.
With premises like that, it seemed that it was only a matter of time for them to encompass the entire IT market. But not so. There are several obstacles, even serious ones.
So much so that ChatGPT itself is not certain of its future, given the difficulties faced by its owner company, OpenAI, which is even in danger of bankruptcy. Let’s find out why.
Table of Contents
Monstrous expenses: maintaining ChatGPT costs a lot of money
The first, major obstacle to the future of the OpenAI company is the cost of maintaining the service. A concept that is as simple as it is unfortunately fundamental. It really costs a lot of money to keep ChatGPT running.
An in-depth analysis by the Indian portal Firstpost showed that OpenAI spends the abnormally high figure of $700,000 per day just to keep Chat GPT online and running. This number, which generates a total expenditure of more than 250 million a year, does not even include the costs of other products, such as the Dall-E image generator or paid intelligences.
The latter (GPT 3.5 and GPT 4), to make matters worse, are still not yielding the desired results. Subscribers to paid professional chatbots are so few that they do not even break even.
This is the main reason why OpenAI is literally burning money with every single prompt entered by users: a loss of as much as $540 million has been reported since the release of ChatGPT. And in this summer, users have also dropped due to increasing competition.
All this, according to a report by Analytics India Magazine, is putting OpenAI in a situation where it may soon declare bankruptcy.
Not just expenses: privacy and moral rules hold back Chat GPT
At least for the moment, OpenAI is sustaining its abnormal expenses thanks to monstrous investments by big American companies. One above all Microsoft with its Bing AI now implemented on all its services.
But the money to be pulled out every day is not the only problem. There are also numerous limits imposed by governments. Which are becoming more restrained with each passing day against the supposed “freedom” of AI.
First, there is a governmental desire from America to make ChatGPT’s complex internal algorithms transparent. Then there is the issue of privacy. Several governments are, obviously and rightly, imposing real roadblocks to the indiscriminate use of private data, prompts and web pages.
Indeed, U.S. authorities have doubts about OpenAI’s work so far, especially with regard to intellectual property and copyright. Numerous artists speak out against the indiscriminate use of their works. Which are given to AIs without any consent or compensation, so that they then know how to copy their style. One example was the portrait app Lensa.
And it doesn’t stop there; moral rules come in as well. As was the case with the criminal AI WormGPT, an uncontrolled AI can produce unsuitable and very dangerous responses. So brakes must be imposed on it, also desired by creator Sam Altman himself. Over time, of course, this limits the expansion of OpenAI, so much so that hackers exploit ChatGPT for scams.
In short, there are many brakes on ChatGPT and its company, which first spearheaded this apparent technological revolution. Whether it will stay that way, we do not yet know. It will also depend on how much Microsoft, Amazon, Google and others like them want to believe and invest in the industry. And, for how many years. In the meantime, eyes open for possible bankruptcies that would change everything.
Read also: Sam Altman, CEO of OpenAI, admits: GPT-4 now scares us, too
