Strategies for Keeping Up with AI Research

The Speed of AI

For anyone that is following the fields of Artificial Intelligence (AI), Deep Learning (DL) or Machine Learning (ML) it can seem that research is speeding past you like a race car sometimes. A quick search on shows that 2,683 new articles related to AI, DL, or ML were announced between April 1st and May 1st of this year (2020). The research community is constantly being flooded with new work (both good and bad) and keeping up can seem like an impossibility. This is only further exacerbated by the coverage that AI is receiving in the media. Whether it is the imminent arrival of self-driving cars, advancements in natural language processing, automated medical diagnosis, or the fear of the cybernetic robot uprising, AI has been getting plenty of attention.

There are, of course, many benefits to the myriad of exposure that AI is receiving; however, for those working in the field, there can also be some negative effects. First, the fire hose of information from new applications, better architectures, emerging sub-fields, etc…, can become a distraction. There is a countless number of rabbit holes to jump down at any moment. While it is important to explore new and emerging solutions that may provide better solutions, if we spend all of our time chasing shiny objects, then we will never make any progress of our own.

Second, the barrage of new research and the growing AI industry can create unwanted anxiety. The pressure to succeed and succeed quickly can shake the confidence of even the most experienced researchers and developers. It can seem like other researchers are publishing twice as fast or companies are developing twice as many applications. This can lead to poor design decisions and wasted time, which undermines our ability to apply AI effectively. Furthermore, this can make us susceptible to group thinking and limit our ability to approach problems with novel and creative solutions.

Ultimately, it is necessary to keep up with the research related to your application area or industry, if you want to stay relevant. The trick is finding the ways that don’t consume all of your time and work best for you. The rest of this article identifies strategies to do just that. It is also important that we allow the state-of-the-art research to inspire us rather than just influence us, but that will be a topic for a future article.

Strategies for Keeping Up

So the good news is there is a plethora of resources available to help us keep up with the research that is relevant to our interests. In this article, I share a number of the resources that I use and discuss some of my strategies for staying current with AI research. The items below are ordered based on my personal priority but everyone learns differently, so I encourage you to try them out and see what works well for you.

0. Be Realistic and Consistent

We all have a limited amount of time to spend on keeping up with AI, so be realistic and consistent. Dedicate a fixed amount of time that fits within your schedule. Recognizing your time limitations should help you to choose the appropriate strategies that will be the most effective for you. Work smart and hard.

1. Read Papers

So as obvious as it may sound it, it is very important to spend some of your time reading papers. This will provide you with the obvious benefit of learning about new and emerging work, but it will also strengthen your ability to distill and understand complex concepts. Reading technical papers is a skill that is crucial for engineering and science, and like all other skills, practice makes perfect. Bonus: critically reading papers is also a great way to come up with new research/algorithm ideas by really challenging everything the authors are presenting.

Here are some of my tips for reading papers.

1A. Stay organized.

So the first tip I have is to keep your papers organized. In my case, I have “ReadMe” folder on my cloud drive that has all of the papers that I would like to read. So, when you have a little bit of time you can go straight to your folder and start reading. Once you have read the paper you can either move it to somewhere else in your digital library or simply delete it.

1B. Read smart.

As I learned quickly during my Ph.D, reading papers can take a lot of time and if you don’t have a good strategy you will never get through all of the papers that you need to. Personally, I have found that a three-phase approach works very well for me. Each phase has a specific purpose and the subsequent stages build on the previous ones. This can greatly speed up the process, because you will often find that getting through phase 2 is sufficient in many cases.

This is my general approach.

  • Phase 1: Read the abstract and the conclusion.
  • Phase 2: Read through the entire paper, but skip any technical details that require lots of mental effort.
  • Phase 3: Read critically and challenge the authors assumptions and assertions.
  • Phase 4 (Bonus!): Code it up! This is obviously more than just reading, but ultimately the best way to truly understand new concepts.

I will be publishing a more detailed article on my approach in the near future. I would also encourage you to read several other strategies and find what works for you.

1C. Set up alerts.

Google Scholar is an amazing resource for many reasons, but the automated alerts are a great way to keep up with what’s new. Alerts can be created based on an author or a search string. Author alerts are great for following the heavy hitters in the field and search string alerts are good for finding new authors.

#artificial-intelligence #resources #productivity #deep-learning #deep learning

What is GEEK

Buddha Community

Strategies for Keeping Up with AI Research
Otho  Hagenes

Otho Hagenes


Making Sales More Efficient: Lead Qualification Using AI

If you were to ask any organization today, you would learn that they are all becoming reliant on Artificial Intelligence Solutions and using AI to digitally transform in order to bring their organizations into the new age. AI is no longer a new concept, instead, with the technological advancements that are being made in the realm of AI, it has become a much-needed business facet.

AI has become easier to use and implement than ever before, and every business is applying AI solutions to their processes. Organizations have begun to base their digital transformation strategies around AI and the way in which they conduct their business. One of these business processes that AI has helped transform is lead qualifications.

#ai-solutions-development #artificial-intelligence #future-of-artificial-intellige #ai #ai-applications #ai-trends #future-of-ai #ai-revolution

Aileen  Jacobs

Aileen Jacobs


Researchers Claim Inconsistent Model Performance In Most ML Research

The process of benchmarking is considered to be one of the most crucial assets for the progress of AI and machine learning research. The benchmark datasets are usually fixed sets of data, which are manually, semi-automatically as well as automatically generated to form a representative sample for these specific tasks to be solved by a model.

Recently, researchers from the Institute for Artificial Intelligence and Decision Support, Vienna claimed that the considerable part of metrics currently used to evaluate classification AI benchmark tasks might be inconsistent. It may result in a poor reflection in the performance of a classifier, especially when used with imbalanced datasets.

For the research, they analysed the present aspect of performance metrics that are based on data covering more than 3500 ML model performance results from a web-based open platform.

#developers corner #ai research benchmark #ai research papers #benchmark #benchmarking ai #bias in ml research #inconsistent benchmark

Murray  Beatty

Murray Beatty


This Week in AI | Rubik's Code

Every week we bring to you the best AI research papers, articles and videos that we have found interesting, cool or simply weird that week.

#ai #this week in ai #ai application #ai news #artificaial inteligance #artificial intelligence #artificial neural networks #deep learning #machine learning #this week in ai

This Week in AI - Issue #22 | Rubik's Code

Every week we bring to you the best AI research papers, articles and videos that we have found interesting, cool or simply weird that week.Have fun!

Research Papers


#ai #this week in ai #ai application #ai news #artificaial inteligance #artificial intelligence #artificial neural networks #deep learning #machine learning #this week in ai

Microsoft Reveals Need To Prioritise Skills To Maximise Value From AI

Microsoft India today released new research revealing that organisations that combine the deployment of AI with skilling initiatives are generating most value from AI. The topline findings of the research underscore that mature AI firms are more confident about the return on AI and skills.

The tech giant recently conducted a global survey with approximately 12,000 people working with enterprise companies. The research surveyed employees and leaders within large enterprises across industry verticals in India, and 19 other countries, to look at the skills needed to thrive as AI becomes increasingly adopted by businesses, as well as the key learnings from early AI adopters.

The survey found a direct link between having the skills needed to thrive in an AI world and the value organisations gain from their AI implementations. The research further reveals that employees are keen to acquire AI relevant skills that are growing in importance and are of value to them personally and to the business. The organisation leaders surveyed predicted that half of all employees will be equipped with AI skills in the next 6-10 years, which is nearly one-and-a-half times more than the present estimations.

#news #ai research for businesses #ai survey #microsoft #microsoft ai for business survey #microsoft ai research #microsoft survey