As I have gone deeper into GenAI and leveraging it to solve real business problems, the transformative potential is obvious but I see there being challenges with widespread adoption (in the short term). Adoption is a topic near and dear to me as it is a proxy for value and GenAI is about as good a PLG (product led growth) solution as any and I had high hopes that it would just get adopted. In fact there is this WSJ article that hints at this - https://www.wsj.com/tech/ai/early-adopters-of-microsofts-ai-bot-wonder-if-its-worth-the-money-2e74e3a2
Here are the challenges I have faced with using it in real life business use cases -
It requires me to communicate in a way I am not used to - I have to provide context, I have to be clear on what I am asking, I have to write the prompt in a specific way. There is a learning curve here.
It is not a - ask a question and you get your answer. There is a fair bit of iteration that needs to happen before you get the answer you are looking for. Reminds me of early days of building Tableau dashboards. You had to iterate and sometimes you know what you want only after you have seen the output and then you change the prompt. Most people are not comfortable with investing the time iterating.
To solve real business problems, requires me to combine multiple steps - set the context, ask a question, provide specific inputs (pdfs / weblinks etc.) review the response, ask follow up questions, now take the output from all the responses and output to word or something. Currently there is too much friction in this process for anyone but the 'innovators' to get excited about this. So selling this broadly is going to be a challenge.
Rate of change is too high. An an example, if I was an early Excel user imagine this - I am still learning how to add 2 numbers in excel and now there are formulas and then there are pivot tables and then there are charts and graphs. I am still at 101 and excel has already released 401. This is the breakneck rate of innovation happening in GenAI. If there is one thing I have learnt, vendors love releasing new features, users (except for the innovators and early adopters) don't like new features as they are already up to their gills with things to do and now they have to learn new stuff.
Most vendors are adding GenAI to their products, but how are they paying for it? The foundational vendors are making money but everyone else building on top of it, cannot pass on that cost (unless you are Microsoft or Salesforce or Google). Not sure how this will play out.
To sum it up, there is tremendous potential for GenAI but to get organizations to adopt and get value is going to take a different approach.