LeoFinance And AI
In a recent post we discussed the belief that, at some point, a great deal of what is generated is going to be done by AI (software).
This led to some very interesting viewpoints.
Where will we be in a 5-10 years regarding this technology. GPT-3 by OpenAI is one of the leading content generation programs out there. We will see how the advanccement takes place but many feel that it is off to a good start.
Of course, there is a downside to this according to many. The idea is that human content would get lost. There simply is no way to distinguish human from computer, at least there will not be with future iterations of this software.
This brings up some interesting ideas regarding Leofinance and Hive.
Training AI With Hive Content
This is an interesting viewpoint. We have a lot of data that was posted on Leofinance over the last few years. Could this be used to train GPT-3 to generate content?
Is that something even worthwhile?
From my perspective, I can see a use case. If we were able to train that to create things for LeoGlossary, that would be outstanding. For example, there are thousands of corporations in the United States alone that could be defined and referenced. This is obviously a mammoth task, something that is too large for one person.
However, if the AI was able to "write" an article about each company, then the workload would be seriously reduced. Even if there was some editing necessary for grammer and corrections, that still would remove a lot of the heavy lifting.
So far, AI of this sort has done well with science and business type papers. The reason is because there are more facts and less opinion. Again, if we look at a company like Home Depot, there are some basics such as CEO, year started, countries operating in, and sector that are beyond dispute.
We also see how our content that was created over the past few years could be something to feed into larger systems. This means that everything we posted on-chain carries with it more importance (this is true for all social media since the AI systems are scouring them all the time).
Moving Up The Rankings
There is an opportunity which could be of extreme benefit to a project like Leofinance.
Much of the ranking within the search engines is a numbers game. The more relevant content, the better.
For this reason, having AI generated articles could be a great entry into the mainstream. Since we are dealing with a rather focused market (non-generalized), we could see some of the concepts starting to carry some weight.
Of course, all this is going to depend upon the quality of the articles. There is no doubt putting together 1,000-1,500 word posts can be rather difficult. However, that is what is required for the likes of Google to really take a site seriously. At the same time, it has to find a lot of infomation about that topic, something that can be a drawback with Leofinance since we are rather diverse in our finance, cryptocurrency, and investing topics.
The refresh of the website offers an opportunity for optimization. Leofinance ranks like crap simply because the site is a mess. Hopefully we will see much better site mapping including a tie-in to LeoGlossary from the main page. This could really make Google a lot happier with the content that is being posted on here.
This goes back to what we discuss fairly often: everything we do on here counts.
As we can see with GPT-3, what is posted on Hive is being used to train some systems. What this returns is still undetermined. However, it can be a grand opportunity for us to really take some of what we are building and project them outward.
Imagine an AI program doing a write up of every earnings report and posting it on Leofinance. That would place a lot of relevant content under the Leofinance heading.
This is just one of the possibilitie. We can see how this could really turn into something powerful.
If you found this article informative, please give an upvote and rehive.
gif by @doze
logo by @st8z
Posted Using LeoFinance Beta
