We have reached the 7th installment of the new series for 2023 related to AI. In just a couple of weeks we are going to venture out of these framing topics into the heart of my journey back into polling and sentiment analysis. Toward the end of that journey it will be very clear on the relationships between AGI, AI, and sentiment analysis. Processing to the point of understanding then adding a degree of directionality in terms of sentiment may be beyond the current state of things at this moment. A lot of debate will go into what is understanding, the occurrence of it, and does it have to be sustained to be relevant. Commonly people try to sum that debate up within the context of next best alternative consideration and that is interesting. However, that debate could be summed up within the analogy of trying to argue that a really well drawn map could provide understanding. It certainly could provide the pathing and the next best routing assuming you have a general idea of where you are going and that path is covered on the map.
Let’s zoom back out for a minute and look at a relatively current webinar. A recent one hour webinar from Stanford online professor Christopher Potts related to GTP-3 & Beyond was pretty interesting. It’s easy to find on YouTube and you can watch it faster than 1x if you enjoy that sort of thing.
Outside of that consideration, I’m wondering if we are going to end up moving from people writing books and articles to a world where people are certifying books and articles for correctness. Some sort of model is used to produce content and then a person edits it to or just verifies its correctness. People have spent a lot of time trying to figure out how to handle natural language processing (NLP). As editors of that content people could act as the ultimate gatekeeper of knowledge and correctness within what is spit out of the systems. We are after all capable of processing natural language ourselves. It’s just a lot harder to explain or rationalize the rule systems that each individual person uses to achieve that objective.
Here are 5 scholarly works related to NLP:
Manning, C., & Schutze, H. (1999). Foundations of statistical natural language processing. MIT press. http://www.cs.cornell.edu/home/llee/papers/topost.pdf
Bird, S., Klein, E., & Loper, E. (2009). Natural language processing with Python: analyzing text with the natural language toolkit. " O'Reilly Media, Inc.".
Chowdhary, K., & Chowdhary, K. R. (2020). Natural language processing. Fundamentals of artificial intelligence, 603-649. https://strathprints.strath.ac.uk/2611/1/strathprints002611.pdf
Hirschberg, J., & Manning, C. D. (2015). Advances in natural language processing. Science, 349(6245), 261-266. https://nlp.stanford.edu/~manning/xyzzy/Hirschberg-Manning-Science-2015.pdf
Manning, C. D., Surdeanu, M., Bauer, J., Finkel, J. R., Bethard, S., & McClosky, D. (2014, June). The Stanford CoreNLP natural language processing toolkit. In Proceedings of 52nd annual meeting of the association for computational linguistics: system demonstrations (pp. 55-60). https://aclanthology.org/P14-5010.pdf
You could go out and watch Christopher Manning’s “Natural language processing with deep learning CS224N/Ling284” 23 video course on YouTube [1]. This one contains a lot of content.
Maybe you were looking for a more media rich introduction to NLP. The team over at TensorFlow also has a collection of 6 videos on the topic that are going to be a lot more visually involved [2]. They are also considerably shorter than the 23 video course from Christopher Manning.
Rounding this one out will be a video from the IBM Cloud team which was pretty decent.
Links and thoughts:
Top 4 Tweets of the week:
Footnotes:
[1]
[2]
What’s next for The Lindahl Letter?
Week 112: Autonomous vehicles
Week 113: Structuring an introduction to AI ethics
Week 114: How does confidential computing work?
Week 115: A literature review of modern polling methodology
Week 116: A literature study of mail polling methodology
If you enjoyed this content, then please take a moment and share it with a friend. If you are new to The Lindahl Letter, then please consider subscribing. New editions arrive every Friday. Thank you and enjoy the year ahead.
Natural language processing