The number of papers are indications of interest and not impact. Indeed more people working on the same problem can generate more ideas. But more ideas do not necessarily generate more impactful ideas when ideas are constrained by groupthink.
What is driving the interest in deep learning is of course its phenomenal success. This leads to more funding and more advanced tools. There are diminishing returns in every field as the low-hanging fruit is picked.
Like in any field the early adopters are always rewarded disproportionately more than the latecomers. Unfortunately, it is human bias to recognize more the pioneers.
To gain the greatest upside requires greater risk. Risk gets magnified when the cost of doing business increases. In Deep Learning, the cost of research is growing exponentially. Hence the greater risk leads to less revolutionary and impactful discoveries.
Furthermore, as a field matures, the majority of innovation is more like engineering where we are optimizing different combinations of known solutions to tackle a new problem. I would argue that most Deep Learning papers are of an engineering nature.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Carlos E. Perez

Carlos E. Perez Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @IntuitMachine

3 Oct
When modern civilization voted away monarchies, we collectively sought to get rid ourselves of leaders that were psychopaths. Yet here we are today.
When optimization is the primary driver of civilization, we structure our lives as if we are cogs in a great machine. As a consequence, our leadership also treats people as if they were also machines.
The fear of AI is because they replace us as cogs in the machinery. Thus we lose our relevance. We cannot stomach the possibility that AI replaces our psychopathic leaders. Thus we lose our agency.
Read 9 tweets
1 Oct
If wealth implies having the luxury of time to engage in your passionate interests, then why don't wealthy intellectuals hire tutors so they can understand complex subjects faster?
The strange thing is consultants are usually hired so that the wealthy do not have to think about some things (for example: financial planning), but it is a rare case where they are hired to do something better.
Yet it is not unusual for the wealthy to hire personal trainers because it's common sense that you can't outsource your physical exercise. So why aren't there personal 'cognitive' trainers?
Read 6 tweets
1 Oct
Finally a credible mathematical framework for understanding how to build deep learning architectures for different problem domains. @mmbronstein
5G's of Geometric Deep Learning: Image
Now every Deep Learning practitioner needs to include groups, geodesics, and gauge invariances in their working vocabulary. Our brains are about to explode!
Read 4 tweets
27 Sep
I don't know if there's a consensus of whether we have no idea how the brain works!
I think many people have failed to express what the brain does at Marr's computational level. You've got to begin with a hypothesis and then frame your evidence in support of the hypothesis. Anything less is stamp collecting.
What we know of the brain is what we know about butterflies if all we do is collect a lot of butterflies. This knowing is not the same as knowing how butterflies generate their wings.
Read 4 tweets
27 Sep
In the context of general intelligence, there is no opposite to the notion of similarity. What we describe as dissimilar is also a kind of similarity.
That's because similarity hinges on a change of reference frame. It is what remains the same when a reference frame is changed.
Dissimilarity is what becomes different when some things remain the same. It's contextual and subordinate to the change in reference frames to determine similarity.
Read 6 tweets
26 Sep
What is so frustrating about the FDA and CDC response to the pandemic is the lack of transparency of their decision-making process. If we are ever to evolve into a world of greater complexity, we need governance to be more transparent in their complex decision-making process.
What is preventing governance from creating a decision-making diagram that shows what information was available for every decision and the sequence of steps leading to the final decision?
As we become more dependent on automated decision-making algorithms, we cannot accept the state where decisions are in a black box and not explainable. If a car crashes without explanation, we should be able to review the decision process to discover the flaw.
Read 6 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!

:(