Hi,
If a someone actually knows what they are doing to the extent that they
can guarantee interesting results at the time of funding i.e. not chasing
yet another random hypothesis, the cost for "seed AI" should not be more
than on the order of USD$5 million in my opinion. If it costs more than
that, then one does not understand the problem as well as one thinks they
do.
Agreed -- unless one has a design that requires a massive amount of
hardware,
which costs more than this. (The Novamente design shouldn't require such
a massive amount of hardware.)
And if you can't raise $5M, there is something wrong with either the
person/people or the idea as presented. Even in the darkest days of the
bust, a really solid idea could *still* raise a million or two. A
late-stage project (the only type that should be funded) could do just
fine on a couple million, since it would be outcome directed spending. By
venture capital standards, the total capital requirements are paltry, even
for my own startup.
Hmmm... I don't agree with this at all.
You seem to be mixing up research with business, to an extent that
makes me uncomfortable.
Creating an AGI is by its nature a high-risk endeavor. And if we work for
3-5 years to create a virtual toddler, planning to then spend N years
teaching and
training it to be an adult seed AI, there's not necessarily a compelling
*business model* associated with this sort of scientific achievement, of the
sort
that appeals to VC's.
I would be curious to get your advice on how to raise business VC investment
for a multi-year AGI project. I have not found VC's to be interested in
this sort
of thing. Rather, insofar as they're interested in AI at all, I have found
them to only
be interested in more narrowly application-
direction AI development. And I have found that pursuing AGI and
application-oriented
narrow-AI simultaneously leads to too many trade-offs, and disappointingly
slow
progress toward AGI compared to what would be possible with a more intensive
AGI
focus.
If you know some VC's who are interested in discussing
explicitly-AGI-focused
business plans involving several years to delivery of any finished product,
please
contact me off-list ;-)
thanks
Ben
-------
To unsubscribe, change your address, or temporarily deactivate your subscription,
please go to http://v2.listbox.com/member/[EMAIL PROTECTED]