Noam Chomsky may have been describing ANN-based models when he said, “There is a notion of success… which I think is novel in the history of science. It interprets success as approximating unanalyzed data
Very interesting, and I look forward to future posts on this. What you're saying resonates with a similar observation about the fields of theoretical physics and cosmology. I'm seeing increased reliance on speculative models, and I worry that we may be drawing conclusions from inaccurate models almost out of desperation due to lack of real progress in the fields.
The notion of hyperreal seems to me to parallel my sense that culture and science are increasingly detached from physical reality. Both seem more and more wrapped up in sheer fantasy.
Yes, I definitely see this in physics and cosmology. I have a master's degree in physics and I like to check in on physics and refresh my understanding from time to time. It took me a while to realize how difficult it is to flesh out the connections between experimental physics and theoretical physics in any kind of detail. How, for example, do you go from the wave equation to experimental predictions? It's all out there, but you have to do a fair bit of work. Ditto general relativity and, say, the orbit of Mercury.
Isn’t all science outside human perception scales subject to hyperreality because we can only create a story based on what the data and math suggest, not the actual reality? We can conceptualize waves and particles but we also know quantum physics, while amazingly useful, is incomplete.
Yes, this is a very relevant and tricky issue. I am not going to attempt to solve this age-old philosophical mystery in a comment, but I can say that pragmatic interconnectedness among our stories and models is possible. The worst parts of hyperreality involve proliferating models that actively discourage exploration of the interstitial zones between models.
We never really talk about 'actual reality' per se, but there is a difference between informal and colloquial descriptions of experiences and models of phenomena. Hyperreality involves giving up on the idea that bridges across models can be built, and can constrain our sense of what is real.
You only put the first part of George Box’s aphorism in your footnote but the second part is more important. Some models are magically excellent in certain domains. The 2008 crash showed what happens when you depend on models outside of their domain to manage risk. The tools that enable this conversation show the power of shut up and calculate. The Dadism, “the right tool for the job” is applicable here.
I think you have to look at the models, in their domain, and stories like one of those Magic Eye pictures and reality comes popping out if you do it right. I was never good at those Magic Eye pictures.
I'd second what Yohan said and only add that our instruments have greatly extended human perception scales. We observe everything from the visible universe down to atoms and their parts.
Observe? We’re measuring things beyond human perception and creating visual renderings of that data, not “observing” them or experiencing them directly through our perception. (I think this is an important distinction causes much confusion when not made.)
Yes, no question. To me measuring *is* observing. The visualization of data can be hugely helpful, but there are other ways to analyze our measurement data. I think that when our data is consistent and repeatable, we're justified in taking it seriously (albeit always contingently).
Consider it a Gaussian with human perception at the peak. Our knowledge of, say, the Solar system down to biological cells is pretty solid. As we move upwards and downwards in scale, our reach is decreasingly less solid. (I think we're still pretty far off the mark with regard to quantum physics and cosmology.)
Noam Chomsky may have been describing ANN-based models when he said, “There is a notion of success… which I think is novel in the history of science. It interprets success as approximating unanalyzed data
Haha wow that's spot on.
Very interesting, and I look forward to future posts on this. What you're saying resonates with a similar observation about the fields of theoretical physics and cosmology. I'm seeing increased reliance on speculative models, and I worry that we may be drawing conclusions from inaccurate models almost out of desperation due to lack of real progress in the fields.
The notion of hyperreal seems to me to parallel my sense that culture and science are increasingly detached from physical reality. Both seem more and more wrapped up in sheer fantasy.
Yes, I definitely see this in physics and cosmology. I have a master's degree in physics and I like to check in on physics and refresh my understanding from time to time. It took me a while to realize how difficult it is to flesh out the connections between experimental physics and theoretical physics in any kind of detail. How, for example, do you go from the wave equation to experimental predictions? It's all out there, but you have to do a fair bit of work. Ditto general relativity and, say, the orbit of Mercury.
Indeed. (The ontology of the wavefunction is *such* a vexing conundrum.)
Isn’t all science outside human perception scales subject to hyperreality because we can only create a story based on what the data and math suggest, not the actual reality? We can conceptualize waves and particles but we also know quantum physics, while amazingly useful, is incomplete.
Yes, this is a very relevant and tricky issue. I am not going to attempt to solve this age-old philosophical mystery in a comment, but I can say that pragmatic interconnectedness among our stories and models is possible. The worst parts of hyperreality involve proliferating models that actively discourage exploration of the interstitial zones between models.
We never really talk about 'actual reality' per se, but there is a difference between informal and colloquial descriptions of experiences and models of phenomena. Hyperreality involves giving up on the idea that bridges across models can be built, and can constrain our sense of what is real.
You only put the first part of George Box’s aphorism in your footnote but the second part is more important. Some models are magically excellent in certain domains. The 2008 crash showed what happens when you depend on models outside of their domain to manage risk. The tools that enable this conversation show the power of shut up and calculate. The Dadism, “the right tool for the job” is applicable here.
I think you have to look at the models, in their domain, and stories like one of those Magic Eye pictures and reality comes popping out if you do it right. I was never good at those Magic Eye pictures.
I'd second what Yohan said and only add that our instruments have greatly extended human perception scales. We observe everything from the visible universe down to atoms and their parts.
Observe? We’re measuring things beyond human perception and creating visual renderings of that data, not “observing” them or experiencing them directly through our perception. (I think this is an important distinction causes much confusion when not made.)
Yes, no question. To me measuring *is* observing. The visualization of data can be hugely helpful, but there are other ways to analyze our measurement data. I think that when our data is consistent and repeatable, we're justified in taking it seriously (albeit always contingently).
Consider it a Gaussian with human perception at the peak. Our knowledge of, say, the Solar system down to biological cells is pretty solid. As we move upwards and downwards in scale, our reach is decreasingly less solid. (I think we're still pretty far off the mark with regard to quantum physics and cosmology.)