#trashofai: Generated Blogs, Books, Podcasts - Content Drop-Shipping

latobibor

Andr谩s T贸th

Posted on December 7, 2023

#trashofai: Generated Blogs, Books, Podcasts - Content Drop-Shipping

[This posts will have typos, weird English phrases and everything associated with real human beings from real cultures writing real text.]

I see article after article on how to generate blog posts, books, "content for the web" by ChatGPT and get on with it.

But let's take a step back, from the How? to question the Why of it.

Would you communicate with your teammate with articles?

In a working situation where you are sitting next to your teammate, most of the time you don't write them long emails.

Because it just doesn't make sense as you have access to the person themself (themselves...? - gender neutral language has its own challenges), you can ask specific questions directly, related to the actual task you are doing.

But instead when we read books, we read documentation, we read emails, we have to filter through a lot of material that might be unrelated to our actual task.

Then this is going to be true:

Pregenerated long texts (e.g. articles, books) generated on topics => lots of unrelated information.
Responses generated to concrete questions => lots of related information.

Generated content: drop-shipping information

I find drop-shipping inherently morally ambiguous: you obfuscate the real vendor, your only value is that you have better marketed yourself. I pay the tax on my oversight, but using your service, I have not gained any value.

When the work of ChatGPT and LLMs are just wrapped, I can't help but feel the same sense of being cheated.

This is the bottom of the barrel of hustle culture: the only value from a transaction is what I gain from it.

We are going to cut out the middle-man (middle-person?)

People will start understanding that they no longer need the drop-shipper and they would learn to skip the middle-man. Only vulnerable people would still fall for it - I would say this is not the industry I want to support.

What can we do instead?

I would say that instead we should work on maybe specialized models, that are very accurate in their own fields, where we spend a lot of time making sure that correct answers are produced.

I hate auto-generated docs, but I would love a bot that answers my question with the relevant snippets from an existing documentation - even if that documentation only exists in its training data.

My message is:

Create value for your customers, don't take it from them!

馃挅 馃挭 馃檯 馃毄
latobibor
Andr谩s T贸th

Posted on December 7, 2023

Join Our Newsletter. No Spam, Only the good stuff.

Sign up to receive the latest update from our blog.

Related

Reasoning 1 - Contextual Conversion
contextualconversion Reasoning 1 - Contextual Conversion

August 4, 2024

CONTEXTUAL CONVERSION
contextualconversion CONTEXTUAL CONVERSION

August 4, 2024

What is AI?
ai What is AI?

July 23, 2017