Re: Will Spiritual Robots Replace Humanity by 2100? -- Conferencenotes + thoughts

Date view Thread view Subject view Author view

From: Eugene Leitl (
Date: Thu Apr 06 2000 - 17:27:38 PDT

Ka-Ping Yee writes:

> Bill Joy: He seems to be proposing the impossible. Not part of
> the "real issue" unless you believe that we can really slow it down.
Social reality model remarkably defective, perhaps not so surprising
for a geek. Why was he scowling, did he expect neoluddism would find
consensus in that particular panel?
> Holland and Koza: They think it isn't going to happen, or that it

Of both Johns, Holland was surprisingly refreshing, considered that
he's the father of whole idea (putting Darwin into the machine in the
'50s and '60s), he is certainly aware of reconfigurable hardware
folks. I could also catch a word with Koza (Beowulf list residents
know he's running a 1 kNode (half a kilonode, actually, considered
that he is using dual-CPU boxes) Beowulf dedicated to GP (since
recently, Java). He's aware of bad shape of the the fitness landscape
when it comes to shotgunning the machine code bits (no fitness-neutral
filaments permeating the high-dimensional machine opcode space), but,
strangely, he sees no merit in making the mutation function part of
the population pool (which in biology is clearly is).

> will take a very long time. I think they're wrong. (I posed the

Probably wrong. I agree.

> question directed at Holland against his main argument that we
> "didn't have enough theory" for even genetic methods to achieve AI.)
> Kevin Kelly: Pundit. Entertaining, but not very substantial.
Hey, give him some slack. He's a sociologist. Out of Control has quite
a few highlights, though.

> Fred Drake: Also on the sidelines. Fun, but didn't talk about
> the "real issue".
It would have been fun if he'd at least mentioned the Aliens'R'Us
angle, considering the convergent evolution of fastest-spreading
autoreplicators (those eventually making contact, when their mutual
growth domains collide). I was very tempted to ask how he would want
to detect communication channels implemented as relativistic material
packet (information redundantly encoded as lattice defects) streams.

> Kurzweil and Moravec: They are thinking about the "real issue".
> Their answer is: we will integrate and augment ourselves with
> the new technology. We have to adapt. It's not a bad thing.
Hans Heil Technohitler" Moravec let his true affiliation slip, when
someone (Bill Joy?) cited a very revealing piece of interview. The man
has actually managed to (briefly) give me a case of
creepy-crawlies. Especially, because I'm alternating between sharing
his viewpoint and opposing it. Here's a guy who wouldn't shed a tear
if the whole monkey business would be resolved in a gordian fashion.

> Merkle: He and Drexler have been thinking about the "real issue"
> for a long time. How do we survive Singularity? Promote serious
> investigation of the issues to direct forward-thinking policy
> decisions.

I found Merkle extremely Telleresque. He should consider addressing
his opponents arguments in depth, instead of just providing shallow
pseudorebuttals. It will certainly lend more credulity to the
case. I've heard several participants reacted annoyed to this level of

> My personal position is that we'll never get it figured out
> unless we have the right tools to create an effective group
> intelligence. Hence Crit, Criticons, Roundup, etc.
> (1997)
> (1998)
> (1999)
> It's about time for my yearly idea now...

If I was trying to make the Net transcend, I'd figure out a way to
mutate machine code robustly (using ALife on mutation functions), and
applied that to distributed neural net simulators plus remote exploit
finders. Of course, creating easily machine-readable content makes the
task easier, once the golem awakens.
> Oh, and Carey: very good of you to take notes, and thank you
> for posting them here! May i forward them to others who were
> not able to attend?

Indeed, it was a valuable service. I wish I had taken time to take notes.


P.S. Btw, I've seen you sitting near Eric W. Forste, but I didn't have
time to say hello, and they you were gone.

Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Thu Apr 06 2000 - 18:00:47 PDT