And for good reasons
too.
As they say : If
you cannot improve upon silence , keep quiet
PROOF :
Yesterday , I asked ChatGPT :
Before
ChatGPT got launched , it got trained on billions of parameters. You must have
" read " , millions of headlines from thousands of online news-papers
. You must have computed the frequency with which , a " Typical "
headline appeared in past 10 years. And , by extrapolating this past data, you
must have, also computed the " Probability
" of similar or almost identical, headline to appear , in tomorrow's
newspapers or after one week or one month
In this
context, I want you to read my following note / logic and then , drawing upon
your massive dataset , predict
" Tomorrow's Headlines Today " / Horoscope & Other (
Equally Fictitious ) Stories
ChatGPT replied :
While it is true that predicting specific future events with absolute certainty is not possible, I can provide
you with a general understanding of how probability and prediction work.
The examples you mentioned, such as a dacoit/murder in the city,
a bus accident, or a political leader getting assassinated, are events that are
influenced by a multitude of complex factors and are difficult to predict accurately.
Predicting future events relies on various methods, including statistical analysis, historical data, trend analysis, and expert insights.
While I do have access to a vast amount of information, including headlines from various
news sources, it is
important to note that my
responses are based on patterns and knowledge learned during training and not
on real-time data.
It's worth mentioning that making predictions about specific events or headlines is not a reliable application of AI.
AI models like ChatGPT are designed to assist and provide
information based on existing knowledge, but they do not have the ability to predict future events or
headlines with certainty.
Astrologers and futurologists, as you mentioned, often use vague and general predictions that could apply to a wide range of individuals and situations.
This allows for interpretation and subjective matching after the fact, making it difficult to objectively assess the
accuracy of their predictions.
In summary, while probability and prediction play a role in forecasting future
events, it is important to approach them with a critical mind-set and understand the limitations of any predictions or forecasts.
Following
is my note which I asked ChatGPT to read :
Horoscope
& Other ( Equally Fictitious ) Stories
Horoscope
& other stories .. ………04 May 2003
Most People live on, on eternal hope of a better tomorrow- even those whose “ past / present ” has been quite good.
So, you cannot blame those whose “ past / present ” is not-so-good, to yearn
for a better tomorrow.
And that makes everyone anxious to
“know” what future holds for them.
This human tendency / weakness is the
“ Fundamental ” on which the business of astrologers & futurologists
thrives.
And what technique do these
forecasters employ to predict people’s future ?
They rely on “ Theory of Probability. ”
They know that “events” recur with
regularity – except that some events occur with greater regularity than some
others.
Some events have a greater
probability of occurrence than some others.
No one may have taken the trouble to
accurately compute the “ frequency of occurrence ( probability) ” of each type
of event, in advance.
But every event is likely to occur in
near future or distant future ( if probability is low).
E.g.
· A
dacoit / murder in the city
· A
bus-accident
· A
boat capsizing etc.
But probability of one such
assassination taking place within next 10 years, goes up to 0.9.
So, there are 3 dimensions to an
event viz ;
· Type
/ Nature of event
· Time
period within which it will happen
· To
whom it will happen / affect.
Now examine :
Ø TYPE / NATURE OF EVENT :-
Practically, these can run into
thousands.
An event can happen tomorrow or it
can happen after one month or after 10 years or a century.
Ø PERSON AFFECTED :-
There are 6 billion humans on this
earth.
Now, if you multiply these 3
dimensions, you have literally TRILLIONS & TRILLIONS of combinations, with
a “
Probability ” attached to each !
This is what , makes “ predicting of future ” so easy !
Something or Else ( event)
is likely to happen to,
· Someone
or Else ( Person)
At
Sometime or other ( Time)
As long as you make your PREDICTION sufficiently VAGUE / GENERAL, in respect of
BECAUSE, WHAT YOU HAVE JUST PREDICTED , IS BOUND TO HAPPEN , SOMETIME OR OTHER TO SOMEONE OR OTHER !
Now if something that is vaguely resembling what was predicted, does actually happen to even 5 % of the people who “ read” your prediction, their confidence / faith in your ability to predict future sky-rockets !
They start believing that you have some God-given, magical powers to predict future ( did it not come true for them ? )
Then these 5 % spread the word (
word-of-month) about your magical powers.
And the remaining 95% for whom, the “
Prediction” did not come true,
Ø Either forget that they ever read the prediction OR
Ø Believe that it was not “ meant” for them , OR
of eternal hope, keeps burning in their hearts– the hope for a better tomorrow ).
But these 95 % never go out and tell anybody that things did not happen as predicted !
They don’t want to look silly !
They don’t want to appear to be
superstitious persons, believing in unscientific, irrational predictions !
So only the good-word spreads.
Only the “positive” stories spread.
So each such story, brings in more
& more person to look-up morning newspaper to see, what future holds
for them.
Forgotten within 5 minutes but
remembered if something ( vaguely) similar did happen by the end of the day.
Unrelated Reading :
OpenAI
boss ‘heartened’ by talks with world leaders over will to contain AI risks 05 May 2023
Extract :
But he said there
is a risk of creating a “super-intelligence that is not really well aligned”
with society’s needs in the coming decade.
He suggested the
formation of a “global
organization, that at the very highest end at the frontier of compute
power and techniques, could have a framework
to license models, to
audit the safety of them, to propose
tests that are required to be passed.”
He compared it to
the IAEA, the international nuclear
agency.
United
Nations Agency for Regulating Artificial Intelligence ( UNARAI ) ………… 30
May 2023
Thank
You , Rishi Sunak : Your Voice will be Heard…………………………………………. 04 June 2023
Extract :
Dear Sam Altman ( sam@open.ai ) /
Satyan Gajwani (satyan.gajwani@timesinternet.in ) :
Like me , millions of Indians , are looking forward to
your live discussions on AI , on 7th June / 6 pm
I am sure, among a host of topics that you will
discuss, how to regulate AI , will be the foremost
The reason, I believe my suggestion for UNARAI scores
over Rishi Sunak’s global AI Watchdog , is that , in UN , we already have a
time-tested forum for participation by 193 nations
With its many “ Agencies “ , UN have demonstrated its
ability to influence / regulate , the conduct of nations
Creating one more “ Agency “ within the overall
framework of UN, will find ready acceptance by all its members
With Regards,
Hemen Parekh
www.hemenparekh.ai /
06 June 2023