Re:Feathered dinosaur tail found in amber (2016)
by tigerlily on Friday December 13, 2019 @06:13AM
(#21780691)
> Hundreds of years from now, any dinosaurs we recreate will be the result of creative genetic engineering and imagination.
Not to mention machine learning. I wonder if we get a bunch of samples from different individuals, could an AI simply just piece it all together?
Re:Feathered dinosaur tail found in amber (2016)
by rvense on Friday December 13, 2019 @06:52AM
(#21780816)
If the information is gone, it is gone.
I'm sure you could devise a "machine learning algorithm" that'll shove some random numbers in there, though.
Re:Feathered dinosaur tail found in amber (2016)
by mc32 on Friday December 13, 2019 @07:26AM
(#21780964)
It'd be like using machine learning to recreate unrecorded speeches of Marcus Aurelius or Alexander the Great word-for-word.
Re:Feathered dinosaur tail found in amber (2016)
by rvense on Friday December 13, 2019 @07:28AM
(#21780973)
And in this climate I'm sure you could get some venture capital for that...
Re:Feathered dinosaur tail found in amber (2016)
by lm28469 on Friday December 13, 2019 @11:46AM
(#21783126)
Given how many AI/ML companies infer meaning from meaningless data the information being gone isn't an issue. /s
Re:Feathered dinosaur tail found in amber (2016)
by overcast on Friday December 13, 2019 @09:25AM
(#21781792)
Machine learning from what? Existing dinosaurs? The data is gone, what you're suggesting is that some type of AI is going to literally create life from nothing. Considering we aren't capable of autonomously driving a car down the road without killing someone, we've got a while to go before that happens.
Re:Feathered dinosaur tail found in amber (2016)
by everdev on Friday December 13, 2019 @10:41AM
(#21782489)
> The data is gone
Correct
> Considering we aren't capable of autonomously driving a car down the road without killing someone, we've got a while to go before that happens.
These two problems are completely unrelated. Also, every form of transportation can result in injury or death, even walking and bike riding, so setting our goal at 0 deaths as a benchmark for success doesn't make sense.
I believe the T-Rex soft tissue [2,3] that has been discovered contained short polypeptide sequences. You can infer the codons from this information, but that elides promotor, signaling, non-coding, etc regions that will be of massive importance. Not to mention you won't get much data in the first place.
Hundreds of years from now, any dinosaurs we recreate will be the result of creative genetic engineering and imagination.
[1] https://www.nature.com/news/dna-has-a-521-year-half-life-1.1...
[2] https://m.phys.org/news/2019-11-mechanisms-soft-tissue-prote...
[3] https://www.smithsonianmag.com/science-nature/dinosaur-shock...