Draw a banana on a piece of paper. Chances are in doing so you subconsciously eliminated the dimension of the banana that is 'non curvy' / 'not really varying as much'. You've mentally done a form of PCA, and eliminated the least interesting dimension in reducing it to 2 dimensions instead of 3.

It is a means of defining a new coordinate system that is ordered by dimensions of decreasing variance.

> A fancy expression to say "using programming to solve economic models"

That fancy expression existed well before the birth of programming.

He was an extraordinary editor. With a creative and talented technical team bubbling up myriad product/software concepts and features, he was able to dictate a highly focused direction based on great instinct, and then sell the hell out of it that focused vision both inside and outside. It's remarkable how few products Apple makes to this day.

While a lot of comments revolve around the efficacy of accurately monitoring for T1/2 diabetes and the past noninvasive trials and tribulations thereof, consider that the market for such a product is very limited by Apple's measure (10's of millions for an addressable market, versus billions for phones/watches). It seems much more interesting if they were chasing a general metabolic health or pre-diabetic application where the annoyance and lethargy of the FDA/medical community can be skipped entirely.

People that think ML/AI isn't statistics typically haven't studied statistics, or have a marketing agenda. I can tolerate the latter as a fact of life. But the former ... there is often a disturbing lack of statistical understanding in "ML/AI" practitioners at the ground level, even though the vast majority of their tooling is built on basic multivariate statistics. It's rather inevitable give the sudden sex-appeal of the field, but will lead to an 'AI winter' as those folks over promise and under-deliver. Computational statistics, statistical learning, machine learning, pick your term certainly continues to progress as computational horsepower improves. But as another commenter noted, physicists/chemists still self-identify with quantum mechanics even though the computational methods/approximations for molecular dynamics continue to rapidly improve.

Not sure about AI winter. I think a lot of advances are fuel by GPUs, ASICs and colossal datasets. Also opensourcing the frameworks makes it easier for new comers.

We can recognize objects, recognize speech, at almost human level accuracies. That's a big milestone when you think about it.

Also technology improves exponentially when a ton of smart people funded by crazy fuck you money work on pushing it forward.

I would second what you said ...