"We find strong evidence of a 'shift' in the importance of application-oriented learning research since 2009 (relative to developments in robotics and symbolic systems research), and that a significant fraction of this upswing in application-oriented learning research was initially led by researchers outside the United States."
The suggested inflection point is 2009. That works on a number of levels. Economically, that's post-Great Recession. Technologically, that's post-smart phone and Amazon Web Services (i.e. cloud computing). Echoing the "Second Machine Age", the way the world worked changed. In conventional understanding, 2009 ushers in a new round of economic restructuring. As I continue through the paper, I will be searching for evidence of this epochal break (episteme in Foucault lingo).
I hope that the abstract itself, particularly the second half, has your full attention.
Bigger than a steam engine: "artificial intelligence also has the potential to change the innovation process itself, with consequences that may be equally profound, and which may, over time, come to dominate the direct effect."
That's an impressive introduction and a big contract to fulfill with the reader. The authors posit a positive Riemann sum, the slope of the innovation slope will change. It has changed. Fair to ask how do they know this to be true? The boldness is gobsmacking. Empirically (albeit anecdotally) they deliver.
I can vouch for the anecdotes. I've seen the same in Loudoun County. What we know is transforming because how we know is different. This is a change on par with the Enlightenment, not the Industrial Revolution.
"But whether or not Atomwise delivers fully on its promise, its technology is representative of the ongoing attempt to develop a new innovation 'playbook', one that leverages large data sets and learning algorithms to engage in precise prediction of biological phenomena in order to guide design effective interventions."
As a result, scientific inquiry itself will change to take advantage of the new tools. Cleveland should take a leadership role in developing such tools. This is the knowledge society, not the knowledge economy.
Take a gander at this sentence:
"while some applications of artificial intelligence will surely constitute lower-cost or higher-quality inputs into many existing production processes (spurring concerns about the potential for large job displacements), others, such as deep learning, hold out the prospect of not only productivity gains across a wide variety of sectors but also changes in the very nature of the innovation process within those domains"
If you will, the knowledge economy application of AI is the more efficient production processes. The knowledge society is the change "in the very nature of the innovation process". The knowledge society is the invention of a method of invention: As the Enlightenment is to the Industrial Revolution; artificial intelligence is to ?
Coming to my mind is Theodore Porter's book, "Trust in Numbers." AI gives social scientists the power they thought they had during the era of logical positivism. We didn't have the right telescope to match our ambition. Now we do.
Betting this part toward the beginning will resonate:
"We focus on the interplay between the degree of generality of application of a new research tool and the role of research tools not simply in enhancing the efficiency of research activity but in creating a new 'playbook' for innovation itself."
A new playbook for innovation itself is Enlightenment 2.0.
I'm five pages in and completely dazzled.
Brilliant: "while developments in robotics have the potential to further displace human labor in the production of many goods and services, innovation in robotics technologies per se has relatively low potential to change the nature of innovation itself."
So the shift noted is network analysis. That's cool in and of itself as a way to identify epochal breaks. Something to keep in mind going forward. Finishing Section I, we are on solid footing for understanding this round of economic restructuring.
The heart of section 2 is the following distinction:
"'GPTs' are usually understood to meet three criteria that distinguish them from other innovations: they have pervasive application across many sectors; they spawn further innovation in application sectors, and they themselves are rapidly improving."
From the literature on general purpose technologies, that's the litmus test. Deep learning, not robotics, passes. Therefore, deep learning can drive economic restructuring.
Another important concept is the "invention of a method of inventing" (IMI). This "economics of research tools" is a useful way to think about knowledge society. GPTs cause breaks in economic history. IMIs cause breaks in knowledge history. To be an IMI, AI should change the paradigm of basic research. That's where "The Book of Why" might help with your conceptualization of knowledge society. Back to the paper:
"The invention of optical lenses in the 17th century had important direct economic impact in applications such as spectacles. But optical lenses in the form of microscopes and telescopes also had enormous and long-lasting indirect effects on the progress of science, technological change, growth, and welfare: by making very small or very distant objects visible for the first time, lenses opened up entirely new domains of inquiry and technological opportunity."
How does AI allow us to see things we've never seen before? My Janelia Research Campus anecdotes answer that question in an affirmative way. A bacterial infection is a general condition that we can now associate with specific organisms and target treatment (i.e. precision medicine). This is a paradigmatic shift in well-being knowledge. The same could be said for identify the specific causes of health disparities.
On a sour note, we are out on own defining the knowledge society:
"Mokyr (2002) points to the profound impact of IMIs that take the form not of tools per se, but innovations in the way research is organized and conducted, such as the invention of the university. GPTs that are themselves IMIs (or vice versa) are particularly complex phenomena, whose dynamics are as yet poorly understood or characterized."
That we will understand these dynamics as no one has done is daunting. Onto section 3 ...
The authors work through the history of AI. Symbolic systems is the AI that has been around for decades. The paper has already covered the other two categories by distinguishing between robotics and neural networks. The break in history:
"in the mid-2000s, a small number of new algorithmic approaches demonstrated the potential to enhance prediction through back propagation through multiple layers"
That's a major breakthrough for the neural networks category of AI, which had been around for about 25 years with disappointing results. This is the birth of AI as a GPT and IMI. Conveniently, they present a matrix:
Deep learning is characterized as both a GPT and an IMI. I should say, it could be a "general purpose IMI". Section 5 attempts to find evidence in support of that bit of speculation. I didn't read through the methodology closely since I'm not concerned with replication. And guess what? That's all section 5 is, a description of the methodology with the raw results.
Section 6 evaluates deep learning as an empirical GPT:
"The first insight is that the overall field of AI has experienced sharp growth since 1990."
"there is a steady increase in the deep learning publications relative to robotics and symbolic systems, particularly after 2009"
"there does seem to be an acceleration of learning-oriented patents in the last few years of the sample"
"By the end of 2015, we estimate that nearly 2/3 of all publications in AI were in fields beyond computer science."
Again, the time line with breaks fits my Foucault model. It also shows the attributes of a GPT. But we are still very early in the game.
Section 7 evaluates the IMI impacts. What's at stake:
"If it is also a general purpose IMI, we would expect it to have an even larger impact on economy-wide innovation, growth, and productivity as dynamics play out—and to trigger even more severe short run disruptions of labor markets and the internal structure of organizations."
The paper goes on to describe these disruptions to what I would now comfortable identify as the knowledge society. This is what you must help the new President of CSU to understand.
Section 8 is the conclusion. Since it is only two paragraphs in length, I will post the second one here:
"Our preliminary analysis highlights a few key ideas that have not been central to the economics and policy discussion so far. First, at least from the perspective of innovation, it is useful to distinguish between the significant and important advances in fields such as robotics from the potential of a general-purpose method of invention based on application of multi-layered neural networks to large amounts of digital data to be an 'invention in the method of invention'. Both the existing qualitative evidence and our preliminary empirical analysis documents a striking shift since 2009 towards deep learning based application-oriented research that is consistent with this possibility. Second, and relatedly, the prospect of a change in the innovation process raises key issues for a range of policy and management areas, ranging from how to evaluate this new type of science to the potential for prediction methods to induce new barriers to entry across a wide range of industries. Proactive analysis of the appropriate private and public policy responses towards these breakthroughs seems like an extremely promising area for future research."
2009 is the temporal break. IMI concerns the knowledge society (the economy of basic science). GPT is the knowledge economy.
