Like it or not, it’s safe to assume AI will be a\u00a0conversation that defines the rest of the 2020s.<\/p>\n
The recent launch of OpenAI’s latest model, GPT-5, comes as ChatGPT ascends\u00a0to the fifth most-visited website<\/a> in the world, ahead of Wikipedia, Reddit, TikTok and Amazon. In just three short years, the software has upended education<\/a>, scrambled the tech labor sector<\/a>, and ushered in something of a pandemic of chatbot-related mental health problems<\/a>.<\/p>\n But for all those technocratic bona fides<\/a>, the software that’s said to be kicking off the next industrial revolution has one major hurdle to clear: it has yet to make any real money.<\/p>\n Despite some terrific promises about GPT-5’s capabilities \u2014 OpenAI CEO Sam Altman went as far as to compare the chatbot update to the Manhattan Project<\/a> last month \u2014 the model’s actual improvements are\u00a0seriously underwhelming<\/a>, putting a damper on our timeline to reach the AI singularity.<\/p>\n While that doesn’t really change much for the average Joe, it does tighten the immense financial pressure facing OpenAI and its peers.<\/p>\n Thanks to the ridiculous cost of running millions of energy-intensive<\/a> AI queries every day, the company is struggling to make a dent in its expenditures, as the vast majority of users remain content to use its free model. Out of 700 million weekly users, OpenAI says only 5 million<\/a> of them actively pay for their services.<\/p>\n Put into monetary terms, in 2024, OpenAI was burning through $2.50 for every dollar<\/a> it brought in. Given that most successful startups take anywhere from 3-5 years to turn a profit<\/a>, that might not necessarily sound like a death knell. After all, they’re going to keep growing, right?<\/p>\n However, current analysis<\/a> suggests that OpenAI \u2014 now six years into its growth strategy<\/a> \u2014 won’t reach a positive cash flow until 2029 at the earliest, and that’s if everything goes according to plan.<\/p>\n That might be well and good in China, where low-cost AI architects like DeepSeek<\/a> are free to prioritize long-term research and development without pressure from third-party investors<\/a>. But in the US, where a few massive tech monopolies rule the roost, time is money \u2014 and firms like OpenAI are spending their investors’ cash like water going through a data center<\/a>.<\/p>\n As\u00a0Reuters\u00a0pointed<\/a> out in a recent analysis, Alphabet, Meta, Amazon and Microsoft \u2014 the latter of which is backing OpenAI \u2014 expect to spend upwards of $400 billion<\/a> to keep the AI dream alive in 2025 alone. The promise of that kind of spending has sent tech shares through the roof, with Microsoft becoming the world’s second $4 trillion company<\/a> this summer.<\/p>\n Those numbers might sound high now, but what they really signal is investor confidence that companies like OpenAI \u2014 currently valued at a hefty $300 billion<\/a> \u2014 will continue to grow in a way never before seen<\/a> in history.<\/p>\n Whether they’ll ultimately succeed depends on the tech itself. If the lackluster release<\/a> of GPT-5\u00a0is any indication, the tech industry has a long climb ahead of it \u2014 and it’s all loose rock from here.<\/p>\n