9.3 C
New York
Tuesday, March 5, 2024

Reflections: 2023

Reflections: 2023

It was a short year. At least it feels like one. Something magical, from a technology point of view happened in the 2nd half — GPT-4, Code interpreter, DALLE-3 and a few other notable announcements leap frogged us to future!

I completed the course from Andrew Ng, on Deep Learning specialization.

This graphic below is how I felt/feel about this course.

Jon Snow, Game of Thrones, MidJourney

Of course, it goes without saying that Andrew is a fantastic educator, and I hope Andrew is nominated/receives the Millennium Technology Award!

If you are interested in doing this course, First get up to speed on your math by reading this book, and follow up the course by a set of videos from Chris McCormick.

Prompt Engineering has become the new programming language to learn. Those interested can read up more about this can click here and here.

Besides tech, my other interests are with Art. We saw a number of big announcements in this space from MidJourney, Stable Diffusion, Open AI’s DALLE and other notable tools (Adobe, Meta, and a number of smaller niche players in this space). This is a rapidly evolving space. Here is an excellent article to read about how these tools are making inroads into Art. If you are curious about all the tools available for artists — see the list here.

The more you know, the less you know. Go for the #blue pill.

As many of you may know, Stable Diffusion is one of the early open sourced diffusion neural networks that propelled this movement. In my self interest to learn more about Stable diffusion, I found this treasure trove of information, which led me to learn more about this topic, which led me to the Coursera (Andrew Ng is the founder!).

I also read a lot of books. I get my books direct from authors who self-publish or from humble bundle, who always gives back to the community.

MidJourney generated wishes

Lastly, people are upset about the uncharted use of data for training our AI. Earlier, we saw a number of artists and authors suing the AI companies as well. This calls for new definitions for “fair use” amongst other things. Regulating this would stifle innovation.

Here are a few of my predictions:

  1. Knowledge graphs as we know them, are done for. Neural networks are the new knowledge graphs.
  2. “Curating” data for enterprise use will not be required to harness knowledge
  3. AI infused tools, will be the norm.
  4. Art with AI will be more widely debated, fought and regulated.
  5. Broad and big forums (UN level) on AI ethics and bias.

As John would say, Onward!

See you in 2024!

Source link

Latest stories