I can’t wait to ask it if Jeff Bezos is actually Dr. Evil 
A “business” chatbot is the last thing we need.
4 Likes
So we all know how well Amazon chatbots work. I don’t have high hopes.
a tier for business users will cost $20 per person per month. A version with additional features for developers and IT workers will cost $25 per person per month.
Q can help people understand the capabilities of AWS and trouble-shoot issues.
So for $20-25 a month, I get to ask it how AWS works, so I can spend $$$ on AWS???
So far it’s sounding like this should be a free Q&A or FAQ guide somewhere.
The tool can automatically make changes to source code so developers have less work to do, Selipsky said.
OH HELL NO, We all see how well Amazon codes as it is, NOPE NOPE NOPE.
The ones on the lower right of most business sites that you THINK will get you to a human but never do are USELESS, so this is going to be worse.
2 Likes
Q possesses immeasurable power over time, space, the laws of physics, and reality itself, being capable of altering it to his whim…
Great, what could possibly go wrong with an “omnipotent” AI that loves practical jokes and being a manipulative ■■■■■■■.
2 Likes
So ‘assholery’ - an official term of the SAS, thank goodness (link) - makes it through the currently-set filters, but the root word ‘■■■■■■■’ ain’t given the same pass?
@Pepper_Thine_Angus, may I ask what the cost in blue cookies will be for this adjustment? 
adjustments made
1 Like
Proof positive that seller support is an AI chatbot
2 Likes
Well then, that turned quicker then we all expected, but did exactly what we all know Amazon AI already does
1 Like
The use of “hallucinations” to describe one of Q’s (maybe all AI’s?) dangers and defect types has me 
AI is as trucked up as the rest of us!
2 Likes
“Hallucination” is the fundamental “problem” (function) of GENERATIVE language.
All of these “chat” AIs aren’t “thinking” about the meaning of the totality of questions posed, they are returning groups of words that have the highest probability of being the words that “should” appear next based on what has already been returned.
Watson that played JEOPARDY analyzed the words of the answer (clue) to create a list of best possible questions. It had to come up with a one or few word question. It was really good at that. But the guy responsible for coming up with the challenge had this to say:
“in the end, it was making linguistic predictions. There’s no really deep interpretation.” What he and others are trying to do now: “Our focus is on not learning what’s the next word or what’s the next number?” he says. “It’s learning the comprehension.”
https://www.axios.com/2021/02/13/ibm-watson-jeopardy-win-language-processing
If “AI” can’t comprehend the real meaning of the question posed to it, how can it possibly “generate” meaningful results? The issue is that the program needs to be able to do more than just glue words together. It needs to be able to determine where it needs to insert fact, maybe do a math calculation, or some other process than just “generate” (spit out) the next most likely word.
3 Likes
I agree, cause that’s all this “AI” is doing today, and that’s not true AI
1 Like
I was so going there if your hadn’t!
1 Like
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.