An AI doesn’t just need context. It needs a rational understanding of the context.
Only context without the connections leads to hallucination in models.
As already mention above.
@sergeliatko Two small notes when we talk about automated workflows in a corporate context:
-
I already read the thread!
So I have invested time.
I don’t use the summary function, because important statements from participants are often lost.
The AIs often focuses on key words and summarizes the superficial intention only.
In my last job in the management of a medium-sized German company, I observed that colleagues used ChatGPT for this or similar purposes to save time.
But both in the company and here in the forum, this leads to misunderstandings.
These AIs should definitely be more context-sensitive. -
I have often read here in the forum that the typing time is measured.
→ “If the proportion of text length to typing time is too high - it’s an AI.”
I write my texts in a translator, I already invest time there so that my contribute is constructive and well thought.
Then I go to copy and paste it into the forum!
(I think several users who are not native English speakers do this.)
The proportion ‘text length / typing time’ is naturally high in this case!
Also, my way of expressing myself is similar to an AI - double dumb.
Procedures that can differentiate more subtly would also be desirable here.
So I’ve typed this text twice - brave new efficient world.
Especially when this distribution of resources is seen in the context of the company and the context of automated workflows.