It would be an understatement to say that keeping up with AI developments is challenging. One of the challenges is to keep away from the daily announcements of new tools and capabilities and focus on underlying changes brought about by the availability and usability of these tools for the average person, whether within the organization or as an individual operating on their personal computer.
I've come to the conclusion that it is professional suicide to wait for employers to provide the tools so you can at least learn what it's all about. The organization will need time to safely integrate these tools within the technology ecosystem. I'm not suggesting anyone bypass their employer's guidance around use of AI, but we should all learn to use AI safely and responsibly, at home and at work.
Most people are likely dabbling with generative AI tools. It takes more than dabbling to become proficient and yield the most benefits. It will take more than dabbling to set up efficient tools within organizations. I am doing some intentional dabbling on my own but would not call myself an expert prompt engineer just yet.
One way to keep up with AI news is to use AI to simplify the task by creating a GPT designed to bring up news about AI -- or any subject for that matter. This is the equivalent of a news aggregator but instead of a list of links, it provides a neat, short summary of the content of the items, and a link (or an indication of the source). I created a GPT that, when prompted, gives me about 5 news items from the past 48 hours related to AI and Knowledge Management. It has led me to some interesting things I would never have come across based on my routine sources of news.
When I prompted the ChatGPT 4 with a similar query, it gave me completely different answers. That is not surprising at all. Even asking ChatGPT the same thing twice would probably yield different answers. The biggest difference is that with the tailored GPT, I am able to identify the main sources of information to focus on and even though it is not always pulling from those resources, the results are much more "on target" than with ChatGPT.
So, the pre-prompting of GPTs is critical in creating more precise boundaries for the data to be used by the AI in the development of an answer. At the same time, it has the advantage of an LLM.
There are also a lot of AI developments that won't affect us directly as employees or individuals outside of work but that we should consider as we design future projects and activities. The breadth of AI-generated data that may become available needs to be accounted for. For example, reading this article, "Using artificial intelligence, better pollution predictions are in the air," it struck me that international development projects will need to plan for enhanced capacity to utilize these new sources of data and project implementors may be able to develop more precise models and Theories of Change based on more sophisticated AI-enhanced models.
In short, we all need to go beyond dabbling in generative AI.