[ExI] Translation AI (was dying?)

Adrian Tymes atymes at gmail.com
Thu May 18 03:53:42 UTC 2017


It is entirely possible to build an organic radio receptor.
http://www.sciencedirect.com/science/article/pii/S1369702106714444 has
a bunch of links for organic circuit elements; it's a short hop from
there to organic radio transceivers.

I've been trying not to spam about CubeCab, which is taking more of my
time these days.  (Most of the day tomorrow, in fact.)  Although, I
can think of one on-topic anecdote from just this past weekend about
the state of translation AI.

We're presenting at the Paris Air Show next month, specifically June
23rd.  I'm not going myself because I've got another presentation the
preceding week and might have another one the weekend of the 23rd
(details likely to be fluid until that weekend), but I am trying to
coordinate logistics.  Part of that is coming up with a pair of
free-standing posters for the booth.

Currently we're thinking one will be all pictures and symbols, and the
other will have text.  PAS asked that the text one be English on one
side and French on the other, and to send them our posters' images for
review.  (The deadline to submit was last Monday.  I assume this was
so anyone who doesn't meet their standards has a month to redo.)  I do
not speak French, but Google Translate does.

The English text came out as kind of a manifesto.  (I can C&P the
English & French text here if anyone asks - but I'll only do it if
asked, to make sure it isn't spam.)  I auto-translated it into French,
translated the result back into English to catch the few sentences
that needed restating (just changed the word choices until the
translation loop came up with something close enough to the original),
and submitted it to PAS for review.

I gather the PAS guys are in France and probably native French
speakers, from their names, email addresses, and general contact
information.  I very strongly suspect they have nontrivial quality
standards: the Paris Air Show is the largest in the world (in the
odd-numbered years its held in; there's another European one in
even-numbered years), with representatives from all the major air
forces and aerospace companies, so the organizers are unlikely to
half-ass stuff.  This is likely to be the strongest test of French
skills that I and any automation I use* will face in the immediate
future.

The simple round-trip translation and correction produced results good
enough for them.  The only cost to me, the business owner needing to
meet foreign marketing standards, was less than an extra half hour of
my time - and now I have (kind of) written a manifesto in French, like
many a proper revolutionary.

* Notice how that comes off, BTW.  It's not "the robots do this for
me", but "automation extends what I can do - and thus what I am
responsible for, so the onus is on me to make sure I use it right".
An increasing amount of this kind of experience leads me to suspect
the Singularity will come via merging man and machine - perhaps
involving mind uploading, or perhaps merely extreme cybernetics
coupled with anti-aging advances - rather than AIs that never were
human replacing humanity.

As such, I wonder if things like learning how to learn (which is
starting to be taught formally), and being more aware of our own
bio-mental architecture, are among the early glimmers of the
Singularity.  What happens if someone discovers how to keep brain
plasticity even in advanced age - and if we figure out how to make
neurons transmit information faster, and that starts seeing wide
adoption?

What does figuring out how to be smarter feel like, to yourself and to others?

On Wed, May 17, 2017 at 4:53 PM, William Flynn Wallace
<foozler83 at gmail.com> wrote:
> Is this group just not interested any more?  Don't want new discussions of
> perhaps old ideas?
>
> Then I"ll post book reviews and health stuff - if no one objects.
>
> To be fair, objectors have to post something themselves.
>
> I can't believe all these brilliant minds don't have anything to say.
>
> Am reading Nate Silver's book Signal and Noise.  Will never believe an
> economist again.
>
> Interesting if absurd idea:  can we build something organic that can receive
> radio signals?
>
> bill w
>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>



More information about the extropy-chat mailing list