Is there an empirical method of measuring progress? How can we account for piecewise progress, for example VR had a massive interest in the 80s, went into a winter in the 90s, reinstalled in 2012 by Palmer Lucky, similarly , AI went into a 10 year winter due to Minsky’s critic of Rosenblatt. It seems that progress is not linear, but stochastic and maybe a complex thing to model, it appears that it is not a monolith of which we arrive to but constantly happening in complex ways.
The perceptron was intended to be a hardware machine, first implemented on software. This theory is similar to the Hardware Lottery[1] published by Sara Hooker , implying that ideas in Software Research are successful not because they are correct but due to the available hardware to solve those problems.
Secondly, what would be necessary for a hypothetical golden age to emerge, is it building a new city, restructuring organisations (university, government), rebirth(renaissance), cataclysm (covid,climate change) or simply moving slowly towards reform.
It seems that lots of contemporary innovations are like this and GDP becomes and ever less reliant way of tracking scientific progress.
If nobody bothered to create a better measure of scientific progress, I would like to create it or to help someone create it or to at least figure out what prevents us from creating it.
Re measuring progress, it’s hard. No one metric captures it. The one that people use if they have to use something is GDP but that has all kinds of problems. In practice, you have to just look at multiple metrics, some which are narrow but easy to measure, and some which are broad aggregates or indices.
Re “piecewise” process, it’s true that progress is not linear! I agree it is stochastic.
Re a golden age, I’m not sure, but see my reply to @BrianTan below re “interventions”.
Is there an empirical method of measuring progress? How can we account for piecewise progress, for example VR had a massive interest in the 80s, went into a winter in the 90s, reinstalled in 2012 by Palmer Lucky, similarly , AI went into a 10 year winter due to Minsky’s critic of Rosenblatt. It seems that progress is not linear, but stochastic and maybe a complex thing to model, it appears that it is not a monolith of which we arrive to but constantly happening in complex ways.
The perceptron was intended to be a hardware machine, first implemented on software. This theory is similar to the Hardware Lottery[1] published by Sara Hooker , implying that ideas in Software Research are successful not because they are correct but due to the available hardware to solve those problems.
Secondly, what would be necessary for a hypothetical golden age to emerge, is it building a new city, restructuring organisations (university, government), rebirth(renaissance), cataclysm (covid,climate change) or simply moving slowly towards reform.
[1] https://hardwarelottery.github.io/
I’m also super interested in this and would love to hear Jason’s thoughts.
As Dietrich Vollrath often points out, technological progress does not necessarily lead to an increase in GDP and sometimes actually lowers it: https://growthecon.wordpress.com/2014/08/25/246/
It seems that lots of contemporary innovations are like this and GDP becomes and ever less reliant way of tracking scientific progress.
If nobody bothered to create a better measure of scientific progress, I would like to create it or to help someone create it or to at least figure out what prevents us from creating it.
I don’t really have great thoughts on metrics, as I indicated to @monadica. Happy to chat about it sometime! It’s a hard problem.
Re measuring progress, it’s hard. No one metric captures it. The one that people use if they have to use something is GDP but that has all kinds of problems. In practice, you have to just look at multiple metrics, some which are narrow but easy to measure, and some which are broad aggregates or indices.
Re “piecewise” process, it’s true that progress is not linear! I agree it is stochastic.
Re a golden age, I’m not sure, but see my reply to @BrianTan below re “interventions”.