Later On

A blog written for those whose interests more or less match mine.

Archive for December 6th, 2019

AI and Economic Productivity: Expect Evolution, Not Revolution

leave a comment »

Jeffrey Funk writes in iEEE Spectrum:

In 2016, London-based DeepMind Technologies, a subsidiary of Alphabet (which is also the parent company of Google), startled industry watchers when it reported that the application of artificial intelligence had reduced the cooling bill at a Google data center by a whopping 40 percent. What’s more, we learned that year, DeepMind was starting to work with the National Grid in the United Kingdom to save energy throughout the country using deep learning to optimize the flow of electricity.

Could AI really slash energy usage so profoundly? In the three years that have passed, I’ve searched for articles on the application of AI to other data centers but find no evidence of important gains. What’s more, DeepMind’s talks with the National Grid about energy have broken down. And the financial results for DeepMind certainly don’t suggest that customers are lining up for its services: For 2018, the company reported losses of US $571 million on revenues of $125 million, up from losses of $366 million in 2017. Last April, The Economist characterized DeepMind’s 2016 announcement as a publicity stunt, quoting one inside source as saying, “[DeepMind just wants] to have some PR so they can claim some value added within Alphabet.”

This episode encouraged me to look more deeply into the economic promise of AI and the rosy projections made by champions of this technology within the financial sector. This investigation was just the latest twist on a long- standing interest of mine. In the early 1980s, I wrote a doctoral dissertation on the economics of robotics and AI, and throughout my career as a professor and technology consultant I have followed the economic projections for AI, including detailed assessments by consulting organizations such as Accenture, PricewaterhouseCoopers International (PwC), and McKinsey.

These analysts have lately been asserting that AI-enabled technologies will dramatically increase economic output. Accenture claims that by 2035 AI will double growth rates for 12 developed countries and increase labor productivity by as much as a third. PwC claims that AI will add $15.7 trillion to the global economy by 2030, while McKinsey projects a $13 trillion boost by that time.

Other forecasts have focused on specific sectors such as retail, energy, education, and manufacturing. In particular, the McKinsey Global Institute assessed the impact of AI on these four sectors in a 2017 report titled Artificial Intelligence: The New Digital Frontier? and did so for a much longer list of sectors in a 2018 report. In the latter, the institute concluded that AI techniques “have the potential to create between $3.5 trillion and $5.8 trillion in value annually across nine business functions in 19 industries. This constitutes about 40 percent of the overall $9.5 trillion to $15.4 trillion annual impact that could potentially be enabled by all analytical techniques.”

Wow. These are big numbers. If true, they create a powerful incentive for companies to pursue AI—with or without help from McKinsey consultants. But are these predictions really valid?

Many of McKinsey’s estimates were made by extrapolating from claims made by various startups. For instance, its prediction of a 10 percent improvement in energy efficiency in the U.K. and elsewhere was based on the purported success of DeepMind and also of Nest Labs, which became part of Google’s hardware division in 2018. In 2017, Nest, which makes a smart thermostat and other intelligent products for the home, lost $621 million on revenues of $726 million. That fact doesn’t mesh with the notion that Nest and similar companies are contributing, or are poised to contribute, hugely to the world economy.

So I decided to investigate more systematically how well such AI startups were doing. I found that many were proving not nearly as valuable to society as all the hype would suggest. This assertion will certainly rub a lot of people the wrong way, the analysts at McKinsey among them. So I’d like to describe here how I reached my much more pessimistic conclusions.

My investigation of Nest Labs expanded into a search for evidence that smart meters in general are leading to large gains in energy efficiency. In 2016, the British government began a coordinated campaign to install smart meters throughout the country by 2020. And since 2010, the U.S. Department of Energy has invested some $4.5 billion installing more than 15 million smart meters throughout the United States. Curiously enough, all that effort has had little observed impact on energy usage. The U.K. government recently revised downward the amount it figures a smart meter will save each household annually, from £26 to just £11. And the cost of smart meters and their installation has risen, warns the U.K.’s National Audit Office. All of this is not good news for startups banking on the notion that smart thermostats, smart home appliances, and smart meters will lead to great energy savings.

Are other kinds of AI startups having a greater positive effect on the economy? Tech sector analyst CB Insights reports that overall venture capital funding in the United States was $115 billion in 2018 [PDF], of which $9.3 billion went to AI startups. While that’s just 8 percent of the total, it’s still a lot of money, indicating that there are many U.S. startups working on AI (although some overstate the role of AI in their business plans to acquire funding).

To probe further, I gathered data on the U.S. AI startups that have received the most funding and looked at which industries they were hoping to disrupt. The reason for focusing on the United States is that it has the longest history of startup success, so it seems likely that its AI startups are more apt to flourish than those in other countries. My intention was to evaluate whether these U.S. startups had succeeded in shaking up various industries and boosting productivity or whether they promise to do so shortly. . .

Continue reading.

Written by LeisureGuy

6 December 2019 at 5:57 pm

Why I like time-restricted eating

leave a comment »

First, though, note that a 10-hour window also works well. But a 4-hour window has the advantage that I don’t have to fool with food after 2:00pm until the next day.

What I like is that it’s simple: don’t eat except between 10:00am and 2:00pm (the 4-hour window I picked). No judgment call required: if it’s food, and if the time is not between 10:00 and 2:00, it doesn’t go into my mouth.

It’s the same appeal that the simplicity of a whole-food plant-based diet has for me: if the food is meat, dairy, eggs, or refined/processed, I don’t eat it. No judgment call required — no need to figure out exactly what “in moderation” means in each particular instance. (Is eating one egg “moderate”? How about two? Three-egg omelets are pretty common — how about a three-egg omelet without the sides? Is that “in moderation”? And so on.)

It’s much easier just not to eat any. That’s simple, and the clarity and ease of making the decision (to eat or not to eat) appeals to me. (In the case of the WFPB diet, it obviously helps a lot that the foods I do eat have tremendous variety and taste good and are filling and satisfying.)

Written by LeisureGuy

6 December 2019 at 5:55 pm

If The Universe Is 13.8 Billion Years Old, How Can We See 46 Billion Light Years Away?

leave a comment »

Ethan Siegel writes in Medium:

There are a few fundamental facts about the Universe — its origin, its history, and what it is today — that are awfully hard to wrap your head around. One of them is the Big Bang, or the idea that the Universe began a certain time ago: 13.8 billion years ago to be precise. That’s the first moment we can describe the Universe as we know it to be today: full of matter and radiation, and the ingredients that would eventually grow into stars, galaxies, planets and human beings. So how far away can we see? You might think, in a Universe limited by the speed of light, that would be 13.8 billion light years: the age of the Universe multiplied by the speed of light. But 13.8 billion light years is far too small to be the right answer. In actuality, we can see for 46 billion light years in all directions, for a total diameter of 92 billion light years.

Why is this? There are three intuitive ways we can choose to think about this problem, but only one of them is right.

1) Stuff is everywhere, and light travels at the speed of light. This is the “default” mode most people have. You can imagine a Universe that’s full of stars and galaxies everywhere we look, and that these stars and galaxies began forming pretty close to the very beginning of everything. Therefore, the longer we wait, the farther we can see, as light travels in a straight line at the speed of light. So after 13.8 billion years, you’d expect to be able to see back almost 13.8 billion light years, subtracting only how long it took stars and galaxies to form after the Big Bang.

2) Stuff is everywhere, light moves at c, and everything can move through space. This adds another layer to the problem; not only is there a ton of stuff that emits light, but those light-emitting objects can move relative to one another. Since they can move up to (but not quite at) the speed of light, by the rules of special relativity, while the light moves towards you at the speed of light, you can imagine seeing twice as far as in the first case. Perhaps the objects now could be as far as 27.6 billion light years away, assuming their light just reaches us now and they speed away from us at almost the speed of light.

3) Stuff is everywhere, light goes at c, stars and galaxies move, and the Universe is expanding. This last layer is the counterintuitive one that most people have the hardest time with. Yes, space is full of matter, which quickly clumps into stars, galaxies and even larger structures. Yes, the light it produces all moves at c, the speed of light in a vacuum. Yes, all of this matter can move through space, mostly due to the mutual gravitational attraction of different overdense and underdense regions on one another. All of that is true, just as it was in the second scenario.

But there’s something extra, too. It’s that space itself is expanding. When you look out at a distant galaxy, and see that galaxy is redder than normal, the common way of thinking about it is that the galaxy is red because it’s moving away from us, and hence the light is shifted to longer (redder) wavelengths the same way a siren moving away from you has its sound shifted to longer wavelengths and lower pitches. But that’s still part of explanation #2; General Relativity adds that extra element in of space expanding.

And as the Universe expands, the fabric of space stretches, and those individual light waves in that space see their wavelengths stretch as well!

You might think it’s impossible to tell these two effects apart. If all you can measure is the wavelength of the light as it reaches your eye, how can you tell whether it’s due to motion or due to the fabric of space? As it turns out, there’s a relationship that exists between the redshift (and hence the wavelength) and the observed brightness of the galaxy, which is a function of distance. In a non-expanding Universe, as we covered earlier, the maximum distance we can observe is twice the age of the Universe in light years: 27.6 billion light years. But in the Universe we have today, we’ve already observed galaxies more distant than that!

So how far can we see in any direction? If the Universe had no dark energy in it at all, the farthest objects . . .

Continue reading. Helpful charts, photos, and diagrams at the link.

Written by LeisureGuy

6 December 2019 at 4:36 pm

Posted in Science

Fine Slant sale ends today, so I used it, along with my last all-natural brush

with 2 comments

This is a Sabini brush, Sabini being the man who was responsible for Rooney brushes. The handle is ebony, so it’s an all-natural-materials brush, and it did a fine job making a lather from Phoenix Artisan’s avocado shaving soap (a limited run from some time back).

I wanted to use the fine slant as a reminder that their sale ends today. It is really a wonderful slant but, as I always note, it requires that you use very light pressure, just enough to keep the razor’s head (barely) touching your skin — so this would not be a good first DE razor for a man accustomed to shaving with a multiblade cartridge. But once he learns good technique and pressure control, this razor is excellent. It’s made of aluminum with a handle design similar to the Fine Marvel razor, but the handle ridges are crisp on this razor, unlike to smoothed-over ridges of the Marvel handle (presumably because the Marvel handle is plated zinc alloy). I like the slant handle much better.

A good splash of Avo Nice Shave aftershave left me feeling good — particularly given the good outcome of my experiment in restricting eating to a 4-hour window. This morning my fasting blood glucose was 5.4 mmol/L (97.2 mg/dL), which is “normal.” And so far I’ve lost 1.3 pounds over the past two days since I started. And, oddly, I don’t feel hungry at all, though I do notice that I have a habit of snacking — and it’s purely a habit: when I stifle the impulse, I realize I’m not actually hungry, just carrying out an habitual behavior. Of course, I am not doing heavy labor. I am pretty sedentary, and hunger pains might be more acute if I were burning more calories.

Written by LeisureGuy

6 December 2019 at 8:02 am

Posted in Daily life, Shaving

<span>%d</span> bloggers like this: