AI Upscaling Remarkable improvement ! USSR Venus pics

@kobin @Russell_Crow @Contrarian

I have been playing with AI Upscaling, I have been using the Soviet Union’s photos of the surface of Venus and been Upscaling the photos, look at the remarkable improvement!!!

*note I had to decrease the size of the AI upscaled photo to 75% due to limitations of Talk.


You get the same effect with politicians. Vague statements get interpreted into what you’d like to hear.


Oh, and I just learned that the soviets landed cameras on Venus in the 1970s.

1 Like

Cool! Looks like a bunch of rocks

1 Like

the soviets landed cameras on Venus in the 1970s

That’s nothing! Back in the '70s the Soviets built a Venus probe/rover. Something malfunctioned, and it fell out of Earth orbit and landed in Wyoming and started running amok. Since it was built for the extreme environment of Venus, it was virtually unstoppable. Luckily there was a secret government office, the OSI, who had a bionically enhanced agent who was able to stop the rover. There was a documentary about the incident on ABC… :smiley:

1 Like

Yes it was called Maximum Overdrive.

1 Like

You’re right. A guess might be wrong, but that’s not productive. What matters is how wrong the guesses are and are those wrongs relevant to the problem? Uncertainty exists, but you can be certain that uncertainty also obeys specific mathematical laws (such as the laws of probability theory) and thus the uncertainty is usually quantifiable in some manner. Machine learning algorithms and statistics work because this is the case.

The optimal guess after conditioning on known information and a fairly flexible model is not the same as a typical guess by a person.

So . . . the two of you are in agreement??

So are you inferring that Skynet may not know best? That sounds like it’s borderline blasphemy.

Given his communication style and the context of the conversation, it is reasonable to conclude that his statement is really about trying to bring the reliability of these methods into question. Otherwise, why would he make such a trivial observation? AI is made to operate in a similar way to human reasoning and human reasoning is overwhelmingly inductive and thus inferential. It is important to set the record straight because there is a significant amount of misinformation and voodoo reasoning surrounding AI. Even if he understands AI well, lay people who read his comment may misconstrue what he is saying. Misunderstandings about AI makes my life more difficult.

1 Like

It’s important to note that Skynet’s problems were about extrapolation rather than interpolation. Nuking the human race was basically unforeseen pathological behavior between its causal reasoning and its higher level decision making processes. Interpolation is significantly easier because we aren’t having to transfer reasoning about one sample from a single population to another sample from another population nor do we have to engage in the counterfactual reasoning needed to make moral decisions. Inferences can be done for extrapolation and are done this way, but it requires a significant amount of effort to make it work and likely wouldn’t apply to this use case.