OpenAI, the developer of internet darling software, ChatGPT, is reportedly heading towards ‘financial issues’.
According to a report by Analytics India Magazine, ChatGPT has seen decreasing traffic over the first half of this year. Internet data analytics site, Similarweb, confirmed the issues highlighting that worldwide visitors to the site dropped 9.7% during the month of June.
Minutes per visit and number of unique visitors also fell by around 8% respectively.
Could it be that the fad is starting to wear off? And what does OpenAI’s roadmap to success or failure look like from here?
Similarweb has been tracking the gradual slowdown of ChatGPT visitor traffic through the year to date, with June marking the biggest drop-off in terms of visitorship.
Firstly, can we rationally explain the consistent decline?
Some are speculating that it’s down to the lack of students across the world handing in work. This would account for the summer holiday months of June and July but January through May should, in theory, be the peak of school assessments and a prime opportunity for AI essay generation.
Others hypothesize that people are either creating their own bots or finding alternatives, many of which are seen to be gaining traction.
The competition
The leading competitor is Google Bard, launched in March of this year. According to AnswerIQ analytics, Bard has seen explosive exponential growth in the months following its release.
At launch, the site received 30.6m visitors, jumping to 49.7m in April and shooting to 142.6m in May. This is not dissimilar to the growth initially seen by ChatGPT after its launch last year.
The difference here is that Bard is second to the party and has had the opportunity to learn and develop from ChatGPT’s market entry product.
The former has a key advantage in that it draws data from the live internet feed which, in theory, will give it a higher chance of being more accurate than ChatGPT, which draws on information last uploaded in 2021 datasets.
ChatGPT also saves its best results for its premium users, who pay $20 a month. The free version is slow, clunky, and more inaccurate. In contrast, there is only one version of Bard (for now) and it is free and much faster than the ChatGPT equivalent.
Bard is also miles ahead of ChatGPT in terms of its user-friendly interface. Not only does it make for easier text copy-paste through nicer formatting, but users can also edit questions after asking them and view multiple responses that it prepares.
Then there’s Character AI, not quite in the realm of direct competition for ChatGPT as a content generator but stealing visits all the same.
It’s got one thing going for it; novelty, The platform, which allows users to receive generated content from celebrities and historical figures, was set up by former Google staffers who wanted a bit of fun with AI.
According to Similarweb, Character.AI is the second most popular stand-alone AI chatbot site and despite worldwide visits dropping 32% month-over-month to June 2023, traffic is still up tremendously from where it was in June 2022.
Show Me The Money
Keeping ChatGPT running is no cheap feat; OpenAI chief executive Sam Altman claimed it ran to the tune of $700,000 a day.
This may be part of the reason why, despite projecting $200m revenue in 2023 and aiming for $1 billion in 2024, OpenAI is apparently operating with escalating losses; somewhere in the region of $540m last year.
Like a trust fund kid trying to make it work, OpenAI appears to be living off the Microsoft moolah from the latter’s purchase last year. But $10bn won’t last forever when operating costs are as high as Altman reckons.
The usual move for young tech companies on a rocketship in public but drowning behind the scenes (and there have been many of them) is an initial public offering (IPO).
Altman’s already scheduled one for another of his ventures, nuclear power startup, Oklo, which is valued at $1bn, so what’s stopping OpenAI from the same path?
The founder recently said that he had no plans for an IPO for OpenAI, based on the odd company structure. Originally started as a non-profit engine, the company now has a ‘capped-profit’ function which means it can raise funds externally as long as the original non-profit operation still remains.
So the very nature of an IPO, to continuously make a lot of people buy in and everyone get super rich, doesn’t fit the mission statement.
So how can OpenAI get more money?
Due to the giant Microsoft buy-out, the beauty is that OpenAI doesn’t need to look for partnerships with third-party or casino sites as some chatbot developers have chosen.
There are plenty of avenues within the Microsoft universe in which OpenAI now exists for it to not just survive, but flourish.
As it retains exclusivity over OpenAI’s abilities, it can integrate the technology as it develops with its hundreds of other functionalities and platforms across research, gaming, products, and developer software.
What’s more, OpenAI functionality could be the kick in the backside of Microsoft’s Bing search engine needs (since its launch, OpenAI has repeatedly seen more monthly visits than Bing). The same goes for the Microsoft commercial suite with 365 and Azure Cloud.
This all could add up to hundreds of millions or even billions of extra revenue for Microsoft if it can channel the technology properly at the right time.
The AI bubble definitely hasn’t burst yet. The sector remains one of the very few start-up areas that Wall St still sees as a viable funding route as the rest of the market reels from high-profile crashes from former unicorns.
But don’t be surprised or alarmed if the initial surge for ChatGPT usage as we know it today; academic assessments, entrance essays, and budget journalism, is a far cry from what eventually becomes of the technology.
With the backing of a company as powerful as Microsoft, ChatGPT will likely be further integrated with our day-to-day lives in ways that we never saw coming.
