GPT & Deep Machine Learning
OpenAI & Microsoft’s ChatGPT took 5 days to acquire 1m users and then grew to 100m users in only 2 months. The next quickest platforms were Spotify with 1m users in 5 months, and TikTok with 100m users in 9 months.
Sources: Statista, https://www.statista.com/ chart/29174/time-to-one-million-users
Andy Kelly
Is this the next human revolution?
Generative Pre-trained Transformers (GPTs), such as OpenAI’s ChatGPT, are a revolution that has arrived. When paired with bespoke interfaces, these large-scale AI language models can deliver ultra-high value at relatively low cost in almost any operation – from writing effective marketing strategies to designing products and providing high-quality service delivery at scale. In short, adoption
of GPT technology offers significant investor value.
GPT language models take several years to develop, require billions of words for training, and are expensive to operate at scale. Despite this, and the known problem of occasionally giving nonsensical answers, the technology is good enough to transform major industries across the world, including Healthcare, Finance and Education.
High-cost creative processes are particularly likely to be transformed because GPT can automate creative tasks, from generating product descriptions to writing code and improving process designs. At present it cannot replace the entire creative workforce because AI models cannot fully replicate originality and intuition – but for how long?
Due the cost of development and operation, and the value opportunity to so many businesses, it is likely that the integration of third-party GPT technology within proprietary operations will become the norm, forcing a different view of third-party risk in the investment thesis.