Editing Talk:Science DAO Framework

== NOTE ==
Click 'Add topic' to separate subjects.<br>
Please sign all comments by typing 4 tildes (~).<br>
:To answer, use colons (:) to indent
::Use two colons (::) to indent twice
:::Etc.<br>
[[User:Craig Calcaterra|Craig Calcaterra]] ([[User talk:Craig Calcaterra|talk]]) 04:26, 27 March 2023 (CDT)
<br>

== How to develop ==
Jonathan & Craig discussion. Parallel development with this page and the [[Science_Publishing_DAO| Science Publishing DAO]] page. We can converge or diverge, then merge. I'll appropriate things from your page, and you from mine. <br>

Make any new page you want by linking to a [[title that doesn't exist]], then click on your new link. [[User:Craig Calcaterra|Craig Calcaterra]] ([[User talk:Craig Calcaterra|talk]]) 19:55, 11 April 2023 (CDT)

== Failures of the scientific academic establishment ==

One of the reasons this SDF is important is that science has been falling behind on crucial developments in engineering. For instance blockchain has no authoritative economic theory that explains its consequences.<br>

As another instance, AI has no rigorous foundation and no authoritative theory around it. They've been working on it pretty seriously since the 60's. But the innovation is so dramatic that it doesn't slow down long enough to do an academic autopsy on the subject. <br>

It's a huge failing of academics, IMO. We are not keeping ahead of these advances that are affecting the public. It used to be, when electricity was unleashed on society, for example, there was a 100 year old physics theory that had explained it all. Similar for nuclear and quantum theory. They preceded the engineering applications, which gave business time to react in more healthy ways. But we're unleashing dramatic changes to society, with AI and P2P communication. But academia is barely even part of the process, and they ignore it until it's settled into the cracks of our lives. But academic analysis is crucial for being able to guide the innovation in healthy ways for society. Our failures to do this, to make sense of these innovations, to stay ahead of them and guide them in positive ways, is a major abdication of our responsibilities as academics. <br>

For instance, math has no theory that explains why NNs solve these statistical problems so well (for instance, overfitting with these weird statistical architectures doesn't seem to be the problem we would expect). Far less do we have any math theory to predict what will happen when we add new functionality, like memory, to these AIs. Far less can math give a model for controlling these AIs on sophisticated problems, like censoring ChatGPT. (Which is pretty crucial if we don't want to re-enact Terminator's Skynet!) We're way, way behind the engineering and even the marketing. So math has failed. Computer science has failed. Econ and law are guilty of their own sins on these subjects. [[User:Craig Calcaterra|Craig Calcaterra]] ([[User talk:Craig Calcaterra|talk]]) 21:57, 26 April 2023 (CDT)
== Supporting information ==
Sabine Hossenfelder discusses her personal experience in scientific academia in Europe.
https://www.youtube.com/watch?v=LKiBlGDfRU8&ab_channel=SabineHossenfelder