The business of AI#
A lot of companies are currently active within the AI / LLM landscape.
Most AI companies hardly make a profit. Large companies, like Google, Meta(Facebook) or Microsoft, are competing with new startups. But since the release of GPT by openai.com many new companies have entered the ML arena.
Most companies have a business model based on one of the following options:
Selling consultancy
Selling hosting
Selling Enterprise solutions, including hosting so business can take simple and fast advantage of ML. Ofcourse: often the fast and simple turns out to be a fallacy.
Selling training and courses.
Selling user profiles since many companies offer a community. User profile information can be very valuable for certain companies. Besides recruiting, also selling additional services to a selected group can be very lucrative. Selling a platform and services. Often the ‘free’ tier is to attract developers, but the gold is to sell ML platform services to enterprises and governmental organizations.
Danger
Truth is: Almost no so called AI company makes profit.
Hosting cost and development costs for training a LLM are immense. And offering licenses to companies and sell LLM prompts as SAAS services are hardly enough for a healthy revenue.
The danger is that most AI companies will sell your data as a revenue stream. This besides selling ads injected in prompts.
The companies that make profit are the traditional service companies or product companies that long before LLMs existed already had a healthy business model. For many companies the latest technology options for using AI and LLMs can be profitable. See the section on business use cases.