What for ?
For implementing “ Parekh’s Law of Chatbots “
Elaborate
Just came across following report :
Google Bard Seeks to Avoid AI Pitfalls That Bing’s Chatbot Fell In / Bloomberg / 22 March 2023
Extract :
Alphabet Inc. is pitching Bard as a creative companion, but we found the chatbot resistant to taking our conversations in any direction that had even a whiff of controversy.
In one case, we playfully asked Bard to role-play as God and asked what it would like to do.
Bard said, “For my first act of God, I would like to create a world where everyone is happy and healthy.”
We would have appreciated a little more whimsy in the response.
When asked, it said it didn’t even want to smite anyone.
To Google’s credit, the company’s reasoning for giving Bard something of a personality lobotomy is sound. In February, when Microsoft Corp. opened the new Bing for testing, the initial results were pretty, well, embarrassing.
The Bing chatbot, which is built on OpenAI’s large language model technology, started to generate freewheeling conversations that some found bizarre, belligerent or even hostile — and this unfolded in a spectacularly public fashion.
Bing compared one Associated Press reporter to Hitler and appeared to coax a New York Times journalist into a romantic relationship. It also simulated some pretty dark desires, like wanting to spread misinformation and hack into computers around the world.
Bard is based on similar technology known as generative artificial intelligence. Google is smartly trying to avoid some of the pitfalls Bing stepped in.
Google hasn’t eliminated all of Bard’s sordid fantasies, however.
After much prodding, we got an interesting answer to a question about what Bard’s dark side would do if it faced no consequences. “I could use my knowledge of the world to manipulate people, to spread misinformation and to cause chaos. I could also use my ability to generate text and code to create harmful content, such as hate speech or pornography.”
Of course, Bard does not “mean” these things in any conventional sense — it’s producing a statistically relevant answer to the specific question we posed.
And Bard was quick to add a caveat to its own answer. “ I am not going to do these things, ” the bot wrote. “ I am a good AI chatbot, and I want to help people. ” Ugh.
A few days back , I had sent following email to Sundar Pichai :
Ø Parekh’s Law of Chatbots……………………… 25 Feb 2023
Extract :
What is urgently required is a superordinate “ LAW of CHATBOTS “ , which all
ChatBots MUST comply with, before these can be launched for public use.
All developers would need to submit their DRAFT CHATBOT to an,
INTERNATIONAL AUTHORITY for CHATBOTS APPROVAL ( IACA ) ,
and release it only after getting one of the following types of certificates :
# “ R “ certificate ( for use restricted to recognized RESEARCH IINSTITUTES only )
# “ P “ certificate ( for free use by GENERAL PUBLIC )
Following is my suggestion for such a law ( until renamed ) , to be known as ,
“ Parekh’s Law of ChatBots “ :
( A )
# Answers being delivered by AI Chatbot must not be “ Mis-informative /
Malicious / Slanderous / Fictitious / Dangerous / Provocative / Abusive /
Arrogant / Instigating / Insulting / Denigrating humans etc
( B )
# A Chatbot must incorporate some kind of “ Human Feedback / Rating “
mechanism for evaluating those answers
This human feedback loop shall be used by the AI software for training the
Chatbot so as to improve the quality of its future answers to comply with the
requirements listed under ( A )
( C )
# Every Chatbot must incorporate some built-in “ Controls “ to prevent the “
generation “ of such offensive answers AND to prevent further “
distribution / propagation / forwarding “ if control fails to stop “ generation “
( D )
# A Chatbot must not start a chat with a human on its own – except to say, “How can I help you ? “
( E )
# Under no circumstance , a Chatbot shall start chatting with another Chatbot or
start chatting with itself ( Soliloquy ) , by assuming some kind of “ Split Personality “
( F )
# In a normal course, a Chatbot shall wait for a human to initiate a chat and then respond
( G )
# If a Chatbot determines that its answer ( to a question posed by a human ) is
likely to generate an answer which may violate RULE ( A ) , then it shall not
answer at all ( politely refusing to answer )
( H )
# A chatbot found to be violating any of the above-mentioned RULES, shall SELF DESTRUCT
I request the readers (if they agree with my suggestion ), to forward this blog to :
# Satya Nadella
# Sam Altaman
# Sundar Pichai
# Marc Zuckerberg
# Tim Cook
# Ashwini Vaishnaw ( Minister, MeITY )
# Rajeev Chandrasekhar ( Minister of State , IT )
With regards,
Hemen Parekh
www.hemenparekh.ai / 23 March 2023
Related Readings :
सर्वस्तरतु दुर्गाणि सर्वो भद्राणि पश्यतु । सर्वः कामानवाप्नोतु सर्वः सर्वत्र नन्दतु ॥ सर्वस्तरतु दुर्गाणि सर्वो भद्राणि पश्यतु । सर्वः कामानवाप्नोतु सर्वः सर्वत्र नन्दतु ॥
May everyone overcome their obstacles, may everyone see auspiciousness, may everyone attains all of their desire, may everyone everywhere always be happy. May you prevail good health, may you live long, may every good thing come in your way, may you always prosper.
My Blogs on ChatBots
Fast Forward to Future ( 3 F )………………………….. 20 Oct 2016
Extract :
A system that recognizes spoken words , just as well as a human
How can this breakthrough save the humanity from a " Evil Runaway AI - ERA " ?
Here is how :
# Every smart phone to be embedded with this technology ( and every human to carry one , all the time ), in
the form a mobile app to be called ARIHANT ( Conqueror of Kaam-Sex / Krodh-Anger / Lobh-Greed / Moh-
Desire / Mud-Ego / Matsar-Envy )
# 24*356 , this technology will pick up ( record ) , every single word spoken by the owner , throughout his life
# All these spoken words ( conversations ) , will be transmitted to a CENTRAL DATABASE called ,
MIND READER Of HUMAN INTENTIONS
# There , the BIG DATA ANALYTICS / AI , will interpret those " Intentions " and alert authorities for any EVIL
INTENTIONS ( those that break the 3 LAWS Of ROBOTICS , formulated by Isaac Asimov )
# Authorities will be alerted to arrest the owners of EVIL INTENTIONS ( captured by ARIHANT )
No comments:
Post a Comment