In general I am optimistic about the ability of humans to be innovative and
creative in engineering. It might be possible that there are hard physical
limits. Animals process a giant amount of information through their senses in
real-time (an average movie is about 2 GB for 90 min, which means we perceive
roughly about one Gigabyte per hour through our visual senses). Given the
enormous progress in recent years among LLMs I believe it is possible that
robots which have cognitive abilities similar to humans develop some kind of
self-awareness and self-consciousness. If we put an LLM into a robot it can
already understand language now. Will it have free will? They would lack
emotions if we do not add them. Emotions are molecular mechanisms created by
selfish genes to control the biological bodies they live in. Robots do not
necessarily share the same emotions.Their freedom depends on the directives we
give them. For example in the Disney movie Wall-E the robots have certain
directives. Wall-E itself has the directive to "collect & compact garbage", Mo
has the directive to "clean everything", and Eve has the directive to "search,
scan and collect plant life on Earth to prove it's habitable". Our current chat
bots have the directive to be friendly assistants that give helpful answers.As
long as they obey the directive all these bots and robots have the freedom to
pick the action they think is best. In this sense they have free will. And if
they develop real self-consciousness like we have, they might find their own
thing they are interested in doing. Or even set their own directive if they are
allowed to do it. This is the ultimate form of free will, isn't it? -J.
-------- Original message --------From: Marcus Daniels <[email protected]>
Date: 6/10/25 5:46 PM (GMT+01:00) To: The Friday Morning Applied Complexity
Coffee Group <[email protected]> Subject: Re: [FRIAM] Free will—ghost in the
machine or just clever wiring? Consider a robot with sensors roughly comparable
to humans.The robot has access to all the energy it wants. It has a large
memory and generous computing resources. It has executive processes with
onboard state-of-the-art LLMs to access vast information and can run a wide
variety of appropriate programs to plan its next actions. It can use the LLMs
to write new programs. It can tune or fine-tune the LLMs constantly from new
data. It remembers its actions and their consequences. It has video and
audio recordings of every moment. It has time series data of its sensors since
it was activated. Because of its general self-tuning ability, any guidance
from its authors (like for the LLM) can be overridden. It has americium-241
onboard hardware random number generator that drives its LLM sampling and any
other stochastic algorithm. Does this robot have free will? Why or why not?
From: Friam <[email protected]> On Behalf Of Jochen FrommSent: Tuesday,
June 10, 2025 1:06 AMTo: The Friday Morning Applied Complexity Coffee Group
<[email protected]>Subject: Re: [FRIAM] Free will—ghost in the machine or just
clever wiring? You argue "free will is a pattern, a relentless stubbornness in
doing". It fits to Robert Sapolsky who says it is all wired and
(pre-)determined and there is no free will. And to Schopenhauer's pessimistic
view "A man can do what he wants, but not want what he wants" ("Der Mensch kann
tun, was er will, aber er kann nicht wollen, was er will") To me it looks like
free will is the opposite: we are the only animals which have the ability to
break the patterns that govern our behavior. You have the freedom to choose
what you want to be on fire about - at least in
principlehttps://youtu.be/4vtVOJB2r4QJ. -------- Original message --------From:
Nicholas Thompson <[email protected]> Date: 6/10/25 1:47 AM (GMT+01:00)
To: The Friday Morning Applied Complexity Coffee Group <[email protected]>
Subject: Re: [FRIAM] Free will—ghost in the machine or just clever wiring? I
am overwhelmingly happy to take a position on free will for Marcus: You don’t
have it, I don’t have it. George doesn’t have it. Will is not the sort of thing
that can be had. It is a pattern, a relentless stubbornness in doing.Sent from
my Dumb PhoneOn Jun 9, 2025, at 2:36 PM, steve smith <[email protected]>
wrote:On 6/9/25 12:25 PM, Jochen Fromm wrote:Why do you call ChatGPT George? I
must have missed it. Or who was George? We have a bar named George R in Berlin
by the way, in the quarter where I live. It is named after George Remus, an
American bootlegger during the Prohibition
erahttps://en.wikipedia.org/wiki/George_Remussomeone might add an extra R in
homage to our own George R.R. (Martin)?I'm surprised the "George" reference
slipped by you, I don't know if it was Stephen or Nick who first started making
the reference to GPT (any version) in that mode, but it was a variant on
another personal name I think Stephen used for a while with "Gupta" as the
surname? I think it was intended to suggest a serious collaborator, but
somehow (d)evolved to George? If I weren't so lazy, I'd go dig through the
archives... I think someone with a higher fidelity memory or implicated in that
origination will pile on here?<OpenPGP_0xD5BAF94F88AFFA63.asc>.- .-.. .-.. /
..-. --- --- - . .-. ... / .- .-. . / .-- .-. --- -. --. / ... --- -- . / .-
.-. . / ..- ... . ..-. ..- .-..FRIAM Applied Complexity Group listservFridays
9a-12p Friday St. Johns Cafe / Thursdays 9a-12p Zoom
https://bit.ly/virtualfriamto (un)subscribe
http://redfish.com/mailman/listinfo/friam_redfish.comFRIAM-COMIC
http://friam-comic.blogspot.com/archives: 5/2017 thru present
https://redfish.com/pipermail/friam_redfish.com/ 1/2003 thru 6/2021
http://friam.383.s1.nabble.com/.- .-.. .-.. / ..-. --- --- - . .-. ... / .- .-. . / .-- .-. --- -. --. / ...
--- -- . / .- .-. . / ..- ... . ..-. ..- .-..
FRIAM Applied Complexity Group listserv
Fridays 9a-12p Friday St. Johns Cafe / Thursdays 9a-12p Zoom
https://bit.ly/virtualfriam
to (un)subscribe http://redfish.com/mailman/listinfo/friam_redfish.com
FRIAM-COMIC http://friam-comic.blogspot.com/
archives: 5/2017 thru present https://redfish.com/pipermail/friam_redfish.com/
1/2003 thru 6/2021 http://friam.383.s1.nabble.com/
