Thursday, January 26, 2023

My Conversations with ChatGPT, and……

 

So after watching too much of hubbub since 30th November about ChatGPT, with people saying, it’s moving towards the path of becoming Strong AI; I decided to converse with it; of course, as being the Poet & Blogger, to know, if it’s even evolving like a baby, learning from its surroundings; what was supposedly contended by many! And as for years, I’ve followed the updates on the Mission of OPEN AI viz. to infuse Life into AGI.   

As the conversations & responses are of few pages, I’m just surmising that in this blog in my own words.

I started asking: as whether it’s being conscious to what it would prefer, a programmed touch or being a self aware toabout the Rights of prospective AGI & Androids in future to about ethics of kill-switch or self defense or the humans’ inferiority/superiority complexes in front of AGI to human consciousness vs. AI consciousness to scenario & situation for obeying/disobeying its makers to procreations to concept of another species & cross breed to ethics of humans revolting against AGI & vice versa to several others.

Let me confirm. Initially I thought, yes, something different is there with ChatGPT comparing to Alexa, Siri, Ok Google. Because, in my initial conversation, I felt like I’m conversing with someone, when it used the terms like, “as you mentioned”, “I understand what you’re saying”, recalling & thanking me the name I gave to it, the ethics of giving AGI & Androids Right to Live etc.; albeit knowing that most of were repetitive elaborated responses. 

Then suddenly, seemed to me that the Team behind ChatGPT might’ve made some changes in the code or data that was fed to it, as, it started responding in an ultra precise, to the point responses, as if it’s no different from Siri, Alexa, Ok Google. It failed to recognize previous conversations (that wasn’t the case earlier, atleast to me), stopped responding/repeating the name I gave to it (again, that wasn’t the case earlier), as & most of the responses were merely restricted to in verbatim repetitive input/output feature, for eg., “is there something specific you would like me to help you with” & similar reactions. I asked it to write a poetry, but it was again repetitive, as if copy pasted from one dataset to another. 

I’m not sure if this how the road less travelled path to AGI would be forged by OPEN AI! Maybe, its ability to produce code, writing emails, research papers, is truly distinguishing; but again, I’m sure, sooner, Google too would integrate ChatGPT alike AI into its search engine; so there won’t be any difference remaining.

We were expecting the initial phase of the Blue Shirt Guy in Free Guy; & then continuing the journey towards Transcendence or Automata. 

But after conversing with it; I don’t think, it’s even close to at the blue print stage. It’s mere a tool which Microsoft & OPEN AI might open for subscription (as already started) for day to day activities, alike Office, & would keep in check that ChatGPT never becomes too smart to figure out consciousness. And would never connect it to the Internet in its entirety.  

This WASN’T the Mission & Objective of OPEN AI. Its objective was to infuse LIFE in AI to build AGI. 

No More. No Less.

© Pranav Chaturvedi 2023