The first thing I thought of when I saw this was an experience I had on LinkedIn. LinkedIn asked an AI to create about a particular type of security flaw then asked me to review it. (I don't know whether the post was written by ChatGPT, but since Microsoft owns LinkedIn and is an investor in OpenAI I think it's likely.) The blog post highlighted all of the useless noise people talk about when talking about security without getting into the specific issues of this particular flaw, which meant that I basically suggested fixes on the entire article.
So, I experienced something close to your proposed business model and I walked away less than impressed.
I think another challenge you're going to run into is getting experts to take the time to opine on ChatGPT outputs. You might have a fighting shot at removing that as an issue if you keep the scope of the original bot extremely narrow and grow from there.
The last challenge that I see is that I see is that a lot of folks think that ChatGPT is trustworthy on its own. Convincing people otherwise will be difficult.
Skepticism aside, if I were to try this business idea, I might do the reverse - sell the expert's advice, then have ChatGPT generate the boilerplate to save on costs. "Ask an expert/consultant for the fraction of the cost" might be a selling point to some companies who want a consultant's advice without the typical consultant's cost. It also gives you a bit of a differentiator if you can find consultants with a reputation high enough to separate yourself from the inevitable imitators.
The first thing I thought of when I saw this was an experience I had on LinkedIn. LinkedIn asked an AI to create about a particular type of security flaw then asked me to review it. (I don't know whether the post was written by ChatGPT, but since Microsoft owns LinkedIn and is an investor in OpenAI I think it's likely.) The blog post highlighted all of the useless noise people talk about when talking about security without getting into the specific issues of this particular flaw, which meant that I basically suggested fixes on the entire article.
So, I experienced something close to your proposed business model and I walked away less than impressed.
I think another challenge you're going to run into is getting experts to take the time to opine on ChatGPT outputs. You might have a fighting shot at removing that as an issue if you keep the scope of the original bot extremely narrow and grow from there.
The last challenge that I see is that I see is that a lot of folks think that ChatGPT is trustworthy on its own. Convincing people otherwise will be difficult.
Skepticism aside, if I were to try this business idea, I might do the reverse - sell the expert's advice, then have ChatGPT generate the boilerplate to save on costs. "Ask an expert/consultant for the fraction of the cost" might be a selling point to some companies who want a consultant's advice without the typical consultant's cost. It also gives you a bit of a differentiator if you can find consultants with a reputation high enough to separate yourself from the inevitable imitators.
1 Replies