Much has been written about the release of OpenAI’s latest product ChatGPT over the last week. Free (for now) it set a new record for the fastest product to attract one million users.
Having only just discussed the potential for generative AI to accelerate the development of strategic foresight last month, an accessible new product has landed. With early access, having already been using DALL-E, OpenAI's generative AI for image creation, I jumped online to test it out.
When asked to "Explain how generative AI might be applied to develop strategic foresight and futures", this is how ChatGPT responded:
"Generative AI is a type of artificial intelligence that is capable of generating new and unique outputs based on a given set of inputs. This can include things like generating new images, text, or other types of data. In the context of strategic foresight and futures, generative AI could be used to help generate new ideas and possibilities for the future.
For example, a generative AI system could be trained on a large dataset of historical data and trends, as well as information about current events and developments in a particular industry or field. The system could then use this information to generate new scenarios and predictions for the future, taking into account a wide range of possible variables and factors.
These predictions and scenarios could then be used by organizations to develop strategic plans and make decisions about the future. By providing a diverse range of possible future outcomes, generative AI could help organizations to be more prepared and flexible in their planning, and better able to respond to a wide range of potential future scenarios.
Overall, generative AI has the potential to be a valuable tool in the field of strategic foresight and futures, by providing organizations with new and unique insights into the future and helping them to develop more effective and agile strategies."
All of which sounds articulate and perfectly plausible, if a little lacking in personality.
On first use ChatGPT appears magical. It even understands some of it’s own limitations, including that because it has only been trained on datasets through to 2021. Ask it anything about a major event in 2022 and it will politely advise that it has nothing to reference.
And it's certainly not infalliable, given that its output is probabalistic, and not based on a genuine understanding of what is right or wrong. It is referencing all manner of opinions, facts and disinformation recorded online. So we all still need to do our homework.
The underlying technology will undoubtedly prove truly disruptive across a broad range of industries. ChatGPT is currently based on a now almost two year old model, GPT3. Next year we can expect the more powerful GPT4 to be released, and without doubt it will develop into a real time application.
Critical thinking, verification and editing will be ever more essential human skills to hone at the start of this new collaborative journey alongside generative AI.