Chapter 19
Ten Substantial Contributions of AI to Society
IN THIS CHAPTER
Working with humans
Solving industrial problems
Developing new technologies
Performing tasks in space
This book helps you understand the history of AI, where it is today, and where it
could go tomorrow. However, a technology is useful only as long as it makes some sort
of substantial contribution to society. Moreover, the contribution must come with
a strong financial incentive, or investors won’t contribute to it. Although the government
may contribute to a technology that it sees as useful for military or other purposes
for a short time, long-term technological health relies on investor support. Consequently,
this chapter focuses on AI components that are useful today, meaning that they’re
making a substantial contribution to society right now.
Some people say that the overpromising of AI benefits today could cause another AI
winter tomorrow (https://codeahoy.com/2017/07/27/ai-winter-is-coming/). In addition, the fear mongering by certain influential people is causing people
to rethink the value of AI (https://www.theinquirer.net/inquirer/news/3013919/elon-musk-spews-more-ai-fear-mongering-is-desperate-bid-for-more-media-attention). Both of these issues are countered by others who feel that the potential for an
AI winter is slim (https://www.technologyreview.com/s/603062/ai-winter-isnt-coming/) and that the fear mongering is misplaced (https://www.cnbc.com/2017/09/21/head-of-google-a-i-slams-fear-mongering-about-the-future-of-a-i.html). Discussion is valuable in assessing any technology, but investors aren’t interested
in words; investors are interested in results. This chapter is about results, demonstrating
that AI has become integrated into society in a significant enough manner that another
AI winter truly is unlikely. Of course, getting rid of the hype so that people can
really understand what AI can do for them would be a plus at this point.
Considering Human-Specific Interactions
People drive sales of products. In addition, people decide what to talk about most,
which creates buzz, which in turn creates sales. Although you probably won’t hear
about the technologies discussed in the following sections on the radio, the level
at which they affect people is amazing. In the first case, an active human foot, people
will actually be able to walk using prosthetics with nearly the same ease as they
walk with a natural foot. Even though the group needing this product is relatively
small, the effects can be widely known. The second and third cases have the potential
for affecting millions, perhaps billions, of people. They’re mundane offerings, but
often the mundane is what becomes expected, which again drives sales. In all three
cases, the technologies won’t work without AI, which means that stopping AI research,
development, and sales is likely to be met with disdain by the people using the technologies.
Devising the active human foot
Prosthetics are big money. They cost a fortune to make and are a necessary item for
anyone missing a limb who wants to have a decent quality of life. Many prosthetics
rely on passive technology, which means that they provide no feedback and don’t automatically
adjust their functionality to accommodate personal needs. All that has changed in
recent years as scientists such as Hugh Herr (https://www.smithsonianmag.com/innovation/future-robotic-legs-180953040/) have created active prosthetics that can simulate the actions of real limbs and
automatically adjust to the person using them. Even though Hugh Herr grabbed major
headlines, you can find active technology in all sorts of prosthetics today, including
knees, arms, and hands.
You may wonder about the potential value of using active over passive prosthetics.
Medical suppliers are already doing the research (see some results in the report at
https://www.rand.org/pubs/research_reports/RR2096.html). It turns out that microprocessor-based prosthetics that rely on an AI to ensure
that the device interacts properly with the user are a huge win. Not only do people
who use active technology prosthetics live longer, but these prosthetics have also reduced direct
and indirect medical costs. For example, a person using an active technology prosthetic
is less likely to fall. Even though the initial cost of an active technology prosthetic
is higher, the costs over time are much smaller.
Performing constant monitoring
Chapter 7 discusses a host of monitoring devices used by medicine to ensure that people get
their medications at the right time and in the correct dosage. In addition, medical
monitoring can help patients receive care faster after a major incident and even predict
when a patient will have a major incident, such as a heart attack. Most of these devices,
especially those that are predictive in nature, rely on an AI of some sort to perform
the work. However, the question of whether these devices provide a financial incentive
for the people creating and using them remains.
Studies are hard to come by, but the study results at https://academic.oup.com/europace/article-abstract/19/9/1493/3605206 show that remote monitoring of heart patients saves considerable medical costs (besides
helping the patient live a happier, longer life). In fact, according to Financial
Times (https://www.ft.com/content/837f904e-9fd4-11e4-9a74-00144feab7de), the use of remote monitoring, even for healthy people, has a significant impact
on medical costs (the article requires a subscription to read). The impact of the
savings is so high that remote monitoring is actually changing how medicine works.
Developing Industrial Solutions
People drive a ton of small sales. However, when you think about an individual’s spending
power, it pales in comparison to what just one organization can spend. The difference
is in quantity. However, investors look at both kinds of sales because both generate
money — lots of it. Industrial solutions affect organizations. They tend to be expensive,
yet industry uses them to increase productivity, efficiency, and most of all, income.
It’s all about the bottom line. The following sections discuss how AI affects the
bottom line of organizations that use the supplied solutions.
Using AI with 3-D printing
3-D printing began as a toy technology that produced some interesting, but not particularly
valuable, results. However, that was before NASA used 3-D printing on the International
Space Station (ISS) to produce tools (https://www.nasa.gov/content/international-space-station-s-3-d-printer). Most people will think that the ISS should have taken all the tools it needs when
it left Earth. Unfortunately, tools get lost or broken. In addition, the ISS simply
doesn’t have enough space to store absolutely every required tool. 3-D printing can
also create spare parts, and the ISS certainly can’t carry a full complement of spare
parts. 3-D printers work the same in microgravity as they do on Earth (https://www.nasa.gov/mission_pages/station/research/experiments/1115.html), so 3-D printing is a technology that scientists can use in precisely the same manner
in both places.
Meanwhile, industry uses 3-D printing to meet all sorts of demands. Adding an AI to
the mix lets the device create an output, see what it has created, and learn from
its mistakes (https://www.digitaltrends.com/cool-tech/ai-build-wants-to-change-the-way-we-build-the-future/). This means that industry will eventually be able to create robots that correct
their own mistakes — at least to an extent, which will reduce mistakes and increase
profits. AI also helps to reduce the risk associated with 3-D printing through products
such as Business Case (https://www.sculpteo.com/blog/2017/08/10/the-artificial-intelligence-for-your-3d-printing-projects-business-case/).
Advancing robot technologies
This book contains a wealth of how robot are being used, from in the home to medicine
to industry. The book also talks about robots in cars, space, and under water. If
you’re getting the idea that robots are a significant driving force behind AI, you’re
right. Robots are becoming a reliable, accessible, and known technology with a visible
presence and a track record of success, which is why so many organizations are investing
in even more advanced robots.
Many existing traditional businesses rely on robots today, which is something many
people may not know. For example, the oil industry relies heavily on robots to search
for new oil sources, perform maintenance, and inspect pipes. In some cases, robots
also make repairs in places that humans can’t easily access; such as in pipes (http://insights.globalspec.com/article/2772/the-growing-role-of-artificial-intelligence-in-oil-and-gas). According to Oil & Gas Monitor, AI makes it possible to interpolate between mining
models, reduce drilling costs, and perform simulations that demonstrate potential
drilling issues (http://www.oilgasmonitor.com/artificial-intelligence-upstream-oil-gas/). Using AI enables engineers to reduce overall risk, which means that oil will also
have a potentially smaller environmental impact because of fewer spills.
The reduced price for oil is part of what has driven the oil industry to adopt AI,
according to Engineering 360 (http://insights.globalspec.com/article/2772/the-growing-role-of-artificial-intelligence-in-oil-and-gas). Because the oil industry is so risk averse, its use of AI makes a good test case
for seeing how other businesses will adopt AI. By reviewing articles on the oil industry,
you realize that the oil industry waited for successes in the healthcare, finance,
and manufacturing industries before making investments of its own. You can expect
to see an uptick in AI adoption as successes in other industries grow.
This book covers all sorts of robotic solutions — some mobile, some not. Part 4 of the book covers robots in general, flying robots (which is what drones truly are
when you think about it), and self-driving, or SD, cars. Generally, robots can make
a profit when they perform a specific kind of task, such as sweeping your floor (the
Roomba) or putting your car together. Likewise, drones are money makers now for defense
contractors and will eventually become profitable for a significant number of civilian
uses as well. Many people predict that the SD car will not only make money but become
extremely popular as well (https://www.forbes.com/sites/oliviergarret/2017/03/03/10-million-self-driving-cars-will-hit-the-road-by-2020-heres-how-to-profit/).
Creating New Technology Environments
Everyone generally looks for new things to buy, which means that businesses need to
come up with new things to sell. AI helps people look for patterns in all sorts of
things. Patterns often show the presence of something new, such as a new element or
a new process for creating something. In the realm of product development, AI’s purpose
is to help discover the new product (as opposed to selling an existing product being
the focus). By reducing the time required to find a new product to sell, AI helps
business improve profits and reduces the cost of research associated with finding
new products. The following section discuss these issues in more detail.
Developing rare new resources
As you can see throughout the book, an AI is especially adept at seeing patterns,
and patterns can indicate all sorts of things, including new elements (the “Finding new elements” section of Chapter 16 talks about this aspect of AI). New elements mean new products, which translate into
product sales. An organization that can come up with a new material has a significant
advantage over the competition. The article at https://virulentwordofmouse.wordpress.com/2010/11/30/an-economic-perspective-on-revolutionary-us-inventions/ tells you about the economic impact of some of the more interesting inventions out
there. Many of these inventions rely on a new process or material that AI can help
find with significant ease.
Seeing what can’t be seen
Human vision doesn’t see the broad spectrum of light that actually exists in nature.
And even with augmentation, humans struggle to think at a very small scale or a very
large scale. Biases keep humans from seeing the unexpected. Sometimes a random pattern
actually has structure, but humans can’t see it. An AI can see what humans can’t see
and then act upon it. For example, when looking for stresses in metal, an AI can see
the potential for fatigue and act upon it. The cost savings can be monumental when
dealing with items such as waveguides, which are used for radio transmission (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4481976/).
Working with AI in Space
Chapter 16 takes you on a tour of what AI can potentially do in space. Even though plans for
performing these tasks are on the drawing board, most of them are government sponsored,
which means that they provide an opportunity that may not necessarily result in a
profit. You also find some business-related research projects in Chapter 16. In this case, the business is actually looking to make a profit but may not be making
one today. The following sections look at space in another way and point to what’s
happening today. AI is currently enabling businesses to earn money working in space,
which gives businesses an incentive to keep investing in both AI and in space-related
projects.
Delivering goods to space stations
Perhaps the greatest AI commercial success story in space so far is the resupply of
the ISS by companies such as SpaceX and Orbital ATK (https://www.nasa.gov/mission_pages/station/structure/launch/overview.html). The organizations make money with each trip, of course, but NASA benefits as well.
In fact, the United States as a whole has enjoyed these benefits from the venture:
- Reduced cost for delivering materials, instead of using vehicles from other countries
to resupply the ISS
- Increased use of U.S.-based facilities such as Kennedy Space Center, which means that
the cost of these facilities is amortized over a long time frame
- Added launch centers for future space flights
- More available payload capacity for satellites and other items
SpaceX and Orbital ATK interact with lots of other businesses. Consequently, even
though only two companies might appear to benefit from this arrangement, many others
benefit as subsidiary partners. The use of AI makes all this possible, and it’s happening
right this second. Companies are earning money from space today, not waiting until
tomorrow, as you might think from news reports. That the earnings come from what is
essentially a mundane delivery service doesn’t make any difference.
Space deliveries are essentially new. Many Internet-based businesses ran at a deficit
for years before becoming profitable. However, SpaceX, at least, appears to be in
a position to possibly earn money after some early losses (https://www.fool.com/investing/2017/02/05/how-profitable-is-spacex-really.aspx). Space-based businesses will take time to ramp up to the same financial impact that
earth-based businesses of the same sort enjoy today.
Exploring other planets
It seems likely that humans will eventually explore and even colonize other planets,
with Mars being the likely first candidate. In fact, 78,000 people have already signed
up for such a trip (see http://newsfeed.time.com/2013/05/09/78000-people-apply-for-one-way-trip-to-mars/). After people get to other worlds, including the moon, many people think the only
way to make money will be through the sale of intellectual property or possibly the
creation of materials that only that particular world will support (https://www.forbes.com/sites/quora/2016/09/26/is-there-a-fortune-to-be-made-on-mars/#68d630ab6e28).
Unfortunately, although some people are making money on this project today, we likely
won’t see any actual profit from our efforts for a while. Still, some companies are
making a profit today by providing the various tools needed to design the trip. Research
does fund the economy.