Bets in AI - Computer Vision

Despite the sometimes exaggerated expectations of the potential applications of AI in everyday life and work in the past 10 years, there are a few elements within the domain that continue to show promise with advances made by researchers and industry fairly regularly. One of those is Computer Vision (CV), which is fundamental to a number of technologies with practical uses, like Augmented Reality. As with applications of other areas of AI, the most practical here is decision support.

Below is a great article on the team and vision at Paige, which is using CV for medical imaging purposes:

https://venturebeat.com/2019/12/18/paige-raises-45-million-to-detect-cancer-with-computer-vision/

Augmented Intelligence

Great write-up that takes a more rational stance on the potential wins from broader adoption of AI / machine learning. I’ve been borrowing the term “augmented intelligence” when describing the (hyped up in recent years) acronym of “AI”. The algorithms that sift through large amounts of historical data, and then try to make suggestions on data that’s input to the model by the end-user — are most effective at suggesting the next step / action for the human — rather than replacing people’s decisions entirely. This will be especially evident in healthcare, where doctor / clinician decisions will never be replaced, but rather only augmented by output from systems.

https://blogs.scientificamerican.com/observations/does-ai-have-a-place-in-medicine/

Learning more about ML and AI

A couple of great, credible sources, from which learn more about machine learning and AI, from folk who have been in the space for years and are deepining the field:

https://ai.google/education/

https://www.coursera.org/learn/machine-learning

https://lms.mitx.mit.edu/courses/course-v1:MITx+6.036+2019_Spring/about

https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-0002-introduction-to-computational-thinking-and-data-science-fall-2016/lecture-videos/lecture-11-introduction-to-machine-learning/

https://openai.com/blog/

Healthcare benefiting from ML

Really exciting to see progress in application of machine learning (ML) techniques to the world of healthcare, in ways that are immediately applicable to improving the ways care is provided. In this case, it should be possible to better know the timing when a strong medical intervention is necessary, and how early.

https://www.csail.mit.edu/news/using-ai-predict-breast-cancer-and-personalize-care

https://www.statnews.com/2019/07/17/an-experimental-ai-system-can-predict-when-pancreatic-cysts-will-become-cancerous/

The machine becomes an advisor that we and our clinicians can start to lean on more, for more peace of mind and a better life.

Brushing up on the latest - staying current

Bit different of a post this time around.

As commonly thought, I also believe it is important for anyone in the technology, research, scientific, or medical related field to continue to brush up their domain knowledge, sharpen their skillset and understanding of their field, ecosystem, and latest techniques. This allows for improved communication with colleagues, clients, collaborators, and improved chances of being able to make educated guesses about what’s next — useful for creating the next innovation, or knowing where to invest capital and people’s time (including your own).

Some current topics that stand out, that I will be looking to review:

https://www.coursera.org/learn/ai-for-everyone

https://www.coursera.org/specializations/jhu-data-science

https://www.coursera.org/browse/computer-science/design-and-product