On 22/11/2014 16:16, Matt Mahoney via AGI wrote:
On Sat, Nov 22, 2014 at 9:24 AM, Tim Tyler via AGI <[email protected]> wrote or
quoted:
Typical software productivity is 10 lines per day or 2000 lines per
person per year. So this is about a 270 person-year effort. A typical
budget at $250K per person-year (Ph.D. level salary plus overhead)
would imply a cost of $67 million. Using my original estimate of 300M
lines for AGI, OpenCog is 0.2% complete.
If we add in some runtime unit tests involving the Wikipedia corpus,
we should be able to get there within a week or so. Maybe beyond
human level - unless you also think that Wikipedia is 90% junk.
No, Wikipedia isn't code. It's training data. [...]
The distinction between "code" and "data" is quite a blurry one, though -
as illustrated by the BASIC "DATA" statement. Computer programs
vary quite dramatically in their "data" content depending on their
application domain. Evolution can create data relatively quickly -
as seen in the example: '/Methinks/ it is /like a weasel/'. The claim
that most of the human genome is more like "code" than it is like
"data" seems to me as though it would need to be supported.
Maybe you think 300M lines is way too high. You could argue that human
DNA is 98% "junk" or non-coding. But it is there for a reason. We have
about the same number of genes (20K) as the roundworm Caenorhabditis
elegans, but it has only 100M bases as opposed to our 3B. Most of its
DNA encodes proteins or RNA. But there is a problem with densely
packed DNA, just like there is with code optimized for size. It is
very hard to modify without breaking something. Humans are more highly
evolved than bacteria eating nematodes because mutations in humans are
more likely to be beneficial and less likely to be fatal. We can
switch genes on and off between coding and non-coding, but C. elegans
can't. It is a lot like a big software project where some of the code
is unused, left over from older versions, but might be useful in a
future version. It is much faster to re-enable it than to delete it
and rewrite it through the slow process of evolution.
The problem here is that DNA archiving is of relatively poor quality. Junk DNA
gets overwritten with LINEs and SINEs - and mutates into oblivion without
stabilizing selection. That's not to say that junk DNA is all totally useless -
but rather that its utility as a reservoir of rarely-used genes is probably low.
In a sexual species, the main reservoir of rarely-used genes is normally going
to be other individuals in the population.
DNA size varies widely between species. The genomes of lungfish and
wheat are huge. But all mammals have at least 2B bases, which suggests
a minimum size for evolution to work.
Maybe. Another hypothesis is a relatively recent common ancestry
from a creature with a fairly large-genome. The distribution of genome
sizes might help to resolve this one - if there's really a wall at the lower
end, that might show up in the distribution of genome sizes.
Maybe you still disagree that 300M lines of code are required. If you
said 100M lines I might agree. If you said 1M lines, then I would have
to ask why after 60 years we are still paying people $70 trillion per
year to do work that machines aren't smart enough to do. A new car
(not self driving) has 30M lines of code. So if AGI were simpler than
a car or an operating system, it should have been solved by now.
Yes: machine intelligence looks like a complex problem - which will take a
little while to solve yet.
The idea that sub-human level machines could reproduce the human explosion in
cultural
evolution seems quite attractive to me, though. Social skills appear to have
been the key
to getting cultural evolution going. Behavioural imitation required the complex
ability to
mentally put yourself in another person's shoes while watching them perform
tasks. For
mammals, this looks as though it may require relatively advanced cognition.
However,
machines can copy each other easily. We can engineer them to be social. I think
this
means that we can forget about human level machine intelligence, work on swarms
of relatively stupid minions and use the magic of collective intelligence and
crowd-
power to get them to perform useful work for us.
--
__________
|im |yler http://timtyler.org/ [email protected] Remove lock to reply.
-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424
Modify Your Subscription:
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657
Powered by Listbox: http://www.listbox.com
