Machine Wisdom

[frnsys]

There are a lot of different English translations of the Tao Te Ching, so I was curious what a computer would produce if it were trained on a bunch of them. The idea was that maybe some commonality of the translations could be distilled this way.

I took 94 translations and trained a simple character RNN for 10 epochs, with some interesting results - example output here

It gets fixated on a few words/phrases - "It is the mother of the world." and variations appear a lot, giving it a weird cult-like vibe

Full source is available here

[enkiv2]

Nice!

On Tue, Nov 24, 2015 at 12:08 PM Francis Tseng notifications@github.com
wrote:

There are a lot of different English translations of the Tao Te Ching, so
I was curious what a computer would produce if it were trained on a bunch
of them. The idea was that maybe some commonality of the translations could
be distilled this way.

I took 94 translations and trained a simple character RNN for 10 epochs,
with some interesting results - example output here
https://github.com/ftzeng/tao/blob/master/machine_tao.txt

It gets fixated on a few words/phrases - "It is the mother of the world."
and variations appear a lot, giving it a weird cult-like vibe

Full source is available here https://github.com/ftzeng/tao

—
Reply to this email directly or view it on GitHub
#171.

[enkiv2]

Nice!

On Tue, Nov 24, 2015 at 12:08 PM Francis Tseng notifications@github.com
wrote:

There are a lot of different English translations of the Tao Te Ching, so
I was curious what a computer would produce if it were trained on a bunch
of them. The idea was that maybe some commonality of the translations could
be distilled this way.

I took 94 translations and trained a simple character RNN for 10 epochs,
with some interesting results - example output here
https://github.com/ftzeng/tao/blob/master/machine_tao.txt

It gets fixated on a few words/phrases - "It is the mother of the world."
and variations appear a lot, giving it a weird cult-like vibe

Full source is available here https://github.com/ftzeng/tao

—
Reply to this email directly or view it on GitHub
#171.