How I keep abreast with AI research in 2023

Read, watch, code, repeat...

Gone are the days when we used to search for libraries and hit the doors just before closure to sneak peak into a new book. The idea of being withdrawn and pensive, sitting alone with the new found friend simply feels nostalgic.

Welcome to the topsy-turvy digital era where information finds you. Flooded with information and hype, AI is a forerunner in this front.

So how do we digest and keep abreast with all the information overload in AI? Here are my ways to keep abreast with the field that finds new fronitiers every month

ArXiV Digests

ArXiv is one of the first places to disseminate AI research findings from the AI research community. Its a ocean of research papers in all branches of science. The only way I find best to tap into that ocean of knowledge is to register on their website, choose fields of interest and subscribe to them. I am subscribed to “Computer Vision and Pattern Recognition” which is my area of interest and expertise. Once subscribed, you receive daily emails everyday as shown below. The emails contain a list of papers, their abstracts and links to pdf copies, if you wish to open and read them.

Everyday I make sure I spend 10 minutes going through the email for the day and pick 1 or 2 interesting papers to read. Moreover, the comments section these days has details on the accepted/submitted details, such as “Accepted to CVPR as oral paper”. I ensure I don’t miss reading at least the abstract of papers accepted in top conferences. With time, you will start smelling the good research from mediocre. You will also build a mental map of top researchers and their expertise.

🎒 Scholar Profiles

While Arxiv ensures every single research article hits your inbox, it leads us to becoming fans of specific people in the field. While mostly for the quality of their work, some sheelry for their writing skills. In addition to it, you would have built a keen interest in a niche within the field. For example, you might be interested in Generative AI and would like to follow top scientists in this specific topic.

More broadly, you may wish to follow pioneers in the field who have spent years and built AI to what it is today! In order not to miss out on any front, Google Scholar is your friend.

Head over to Google Scholar and search for your favourite scientists and professors and simply start following them. You can choose to receive alerts whenever they upload a new article. You will get email notifications for every paper or patent they publish.

From the menu on the top left of the scholar page (shown above), you can even save articles to your library to read later. If you are unsure who to follow, I am about to publish how to find and follow top researchers in AI. Please stay tuned.

If you are a researcher in the field, I encourage you to create a scholar profile. Even if you just have a thesis and not a research paper, Google Scholar allows you to upload it. And in my opinion, Google Scholar is the best place to showcase your work in the field. Its a great feeling to see an article against your name.

Conferences

Even with the fancy remote working model playing productivity charms, there is nothing like in person contact.  If you can afford the conference and the travel cost, there is nothing like attending these and bumping into top scientists and talking to them about their work.

These days with AI becoming highly commercial, top companies are choosing to hire talent from conferences. Plus, companies are even posting job adverts in conference websites. Take for instance, the top Computer Vision conference which is CVPR. It has a separate page dedicated to job postings.

If you wish to learn more about top conferences in AI, please watch out for my upcoming posts.

📰 Newsletters

With the advent of substack, there is no shortage of newsletters in the AI space. From my experience of repeated subscription and unsubscription, I have narrowed down to two nesletters that are quite useful for me.

• For my quest for AI research related news, The Batch from DeepLearning.ai works best
• For my itching to code and get my hands dirty, towardsdatascience newsletters work best. I am also inclined to mention about hackernoon news articles if you wish to gain a more broader picture of the happenings.

Note that I am not sponsored by any of the above companies.

📹 Social Media

Twitter, Facebook and more recently YouTube seem to be the most happening places for AI news.

Twitter. I follow the official Twitter accounts of top AI labs such as Google Research, Meta AI.

Furthermore, I also follow the twitter accounts of top conferences such as ICLR, CVPR, ICML. Some of these accounts seem to be dedicated to just a single year of the conference. Feel free to checkout accounts I follow on Twitter to find if its works for you.

Facebook. There are a couple of groups quite active on Facebook in sharing AI resources. They are the deep learning group and AI, ML and DL group

YouTube. I am a huge fan of learning with all senses involved and YouTube is emerging to be one of the best free learning platforms for audio visual learning. I am more inclined towards subscribing to channels that are run by experts or practisioners in the field rather than commercially driven channels that focus on views. I am reluctant to direct you towards one or two channels as I have a handful of friends who are youtube content creators. I don’t want to favour one over the other.

😑 By the way, why not subscribe to our channel AI Bites and keep up to date with AI research papers, ideas and even podcasts to come. 😺

Conclusion

Every human is different and have different learning styles. The above researces are some that work best for me. You may find constant bombardments of emails spamming your inbox rather than enriching your knowledge. If you find any other resources useful, please post a comment so that its useful for other researchers.

Thank you and see you in my  next …