Evolution Continues: Will computers and humans be inseparable?

Judy cries in her crib not long after birth. A device called a “baby tender” on her crib recognizes the cry. It begins playing soothing music. It signals the mother’s smartphone. She acknowledges the alert but must finish a task before coming. She speaks through the device to the baby: “I’m coming, sweetheart. Be there soon.” The baby calms.

As Judy gets older, many of her toys have built-in computer chips and are connected wirelessly to the Internet. A toy bear may talk with her, maintaining a conversation by running an application supported in the cloud. The screen on its tummy can display pictures to support or motivate the conversation.

Even before she begins school, Judy gets her own smartphone. She converses with the digital assistant built into the smartphone to hear information or make connections. She learns to read early because of a reading game that can speak words on the screen and correct mistakes when she reads out loud. Her access to information is expanded well beyond innate human capabilities even before she can read and write; the assistant retains information she wants to retain with accuracy and quantity beyond the reach of human memory. Judy can text or speak to get help from the assistant.

The same digital assistant is available through several devices, ranging from personal computers, smartphones and smartwatches to home speakers and automobiles. It becomes increasingly personalized as she uses it. Her first instinct if she needs to know something is to ask the assistant. The assistant understands human language directed at it. It can engage in conversation to clarify a request when it doesn’t fully understand.

The assistant can also be proactive if it thinks it perceives a need based on either a current conversation, a calendar entry, or an incoming message. Judy comes to depend on the assistant to label messages by priority and block spam. In college, it helps with studies and information sources.

The assistant is provided by a large company and periodically augmented with increasing accuracy and features. Judy can pay for enhanced services customized to her interests. Entrepreneurs develop such enhanced services as a financial opportunity.

Judy often buys items through her digital assistant, using it much like web search. Online sales outlets deliver her requested items.

She controls connected devices in her home with voice commands. She may change the target temperature of a thermostat, turn on a light, or speak to someone ringing the house’s doorbell.

The assistant also becomes technical support as computer applications become increasingly complex and constantly change through automated software updates. She can ask how to do certain things with a piece of software in both personal and work environments. If the digital assistant can’t interact with the software directly through an Application Programming Interface provided by the software developer, it can walk Judy through the steps to reach her goal. Digital assistants as interfaces to other software become almost a requirement as the number of features in software packages and web services grow and become more complex.

At work, the digital assistant has access to work-related tools. Enterprise software can be addressed by human language. The company’s human resources department has a chatbot that answers typical questions.

A child who grows up with such experiences would almost view the connection with computer intelligence as part of them. Judy doesn’t think twice about using shoes that make her able to walk on rough surfaces or hot pavement; shoes extend her humanity beyond the limitation of bare feet. Similarly, her personal assistant’s help seems a natural part of being a modern human. Arguably, connecting to computer intelligence through human language is the most direct connection to a machine possible without a brain implant.

If the assistant is hacked or malfunctions, it is almost like contracting a physical illness. Until the assistant is cured, Judy is disabled. A major and critical feature of the assistant functionality is thus backup of data and cybersecurity. Today, we panic if we can’t find our smartphone. In the future, a failure of a personal assistant is almost like losing a limb — it is part of an individual. There is also an emotional connection, much like the affection we develop for a pet, but perhaps more like our attachment to a close friend.

There is no technological hurdle to making this scenario a practical future. Everything described is available today with some limitations. Personal assistants such as Amazon’s Alexa, Google Assistant, and Apple Siri are extensively used today. These large companies are investing heavily in making the “personal assistants” understand speech better and provide crisp answers to questions when possible.

The general assistants have ways that the user can reach other companies’ conversational applications: Google calls the conversational apps from other companies “actions,” Amazon calls them “skills,” and Apple “shortcuts.” In early 2021, Amazon indicated that there were over 100,000 unique Alexa skills globally.

If the trend this article suggests is valid, every company will want to be able to connect with your indispensable personal assistant. Such a connection will be as critical as having a company web site. The business opportunity is undeniable; financial incentives will motivate providers of the technology to improve performance and add features to the digital assistant portals over time, as well as creating new competitors. The personal assistant portals will continue investing in tools making it easier for outside companies to develop digital assistants reachable through their general assistant in their fight to be your portal of choice.

If Judy needs help in learning in school, her digital assistant can help. In general, digital assistants can help level the economic playing field for those whose ability level or early family disadvantages make learning more difficult. One way is they help is to develop talents other than academic through introducing hobbies that leverage other aptitudes.

This leveling of the academic playing field is arguably a major societal issue. Recent books by Michael Sandel (The Tyranny of Merit) and Fredrik DeBoer (The Cult Of Smart) emphasize that opinion-makers (let’s call them the “elite”) — almost all college graduates — look down on those who haven’t graduated college, some 60% of the US population. Both authors argue that the elite assume their only duty is to assure “equal opportunity” to get an advanced education. If an individual can’t get into or complete college, it is supposedly their fault since they had a “fair chance.”

Both Sandel and DeBoer point out that not everyone is capable by their intrinsic abilities or economic situation to complete a college education. Sandel points out that the quiet and often unconscious assumption of inferiority of those without a college degree by the (typically wealthier) elite leads to a sense of humiliation and anger by those who don’t have one. Arguably, this issue is one major reason for a current populist trend where those feeling dismissed and disadvantaged find something to blame — something to justify their anger. Sandel and DeBoer characterize what this article is calling “the elite” somewhat differently, but the concern over the tyranny of smart is similar.

Trends creating economic classes with differing interests have a major impact. They lead to disparities in income and assets. Many economists discuss this societal dichotomy in terms of growing economic inequality. Without specifically citing education, economists such as Thomas Pikkety simply note that in today’s economy, capital usually grows faster than income. The rich get richer, and the poor stay poor. This economic trend exacerbates the impact of the tyranny of smart. Getting into the club of smart with a college degree may be necessary to have sufficient income to grow your capital.

Digital assistants may help reduce the difference in individual’s core capabilities by enhancing them through a close connection to computer intelligence. Almost all of us today use computer intelligence to augment our human intelligence. Web search is the most obvious example. Word processing software checks our spelling and grammar. GPS map guidance helps us get from one place to another. Often it isn’t obvious computers are helping; most thermostats today use small computers to maintain a temperature we set. When we drive a modern car, our steering wheel doesn’t turn the wheels and stepping on the brake pedal doesn’t press the brake against the wheel; these actions instruct microprocessors to pass on the command to digitally controlled actuators. This is so intrinsic to automobile operation that a shortage of chips significantly slowed auto production lines recently.

In the near future, almost every child will have a mobile device and a digital assistant early in life irrespective of their economic situation. Even if some parents hesitate to provide devices to their children, suppliers of the general personal assistants will be motivated competitively to insure broad early adoption. The philosophy “no child left behind” might even motivate a government program providing physical devices with digital assistants and educational software for low-income families.

There will of course be resistance and concern that digital assistants are reducing the parental interaction with the child, damaging the child’s emotional growth and social skills. The same argument could be made about television or game systems, but that hasn’t prevented those systems from being available to most children. It will remain a parental responsibility to encourage social interaction.

The expanding resources provided by computer-based systems motivate the increasing use of human language to connect more easily with those resources. The Language User Interface (LUI) option is increasingly important as the Graphical User Interface (GUI) that has long provided intuitive access to computer intelligence becomes increasingly overburdened with too many features and too many applications. The GUI also faces issues with small smartphone screens limiting viewing space and big fingers being a poor pointing device. Too many options on menus, too many clicks, and too much typing to complete a task will compete increasingly with the option of “just say what you want.”

The historical importance of the GUI in bringing computer intelligence to the masses is an instructive analogy. The GUI slowly became ubiquitous once its advantages in using human instinctive visual abilities and understanding of the desktop metaphor and pointing became obvious. The LUI is likely to see a similar universal adoption over time as an alternative and supplement to the GUI.

The trend toward digital assistants becoming an indispensable portal to computer intelligence has its downsides, particularly if it becomes an essential companion as this article suggests. Issues include a potential loss of privacy and the control of content by governments or companies.

To say the assistant is “personalized” is to say it has access to information specific to the user, that a user must give up some privacy. It will learn which type of stories a child prefers. The assistant can only “call Sally” if provided with Sally’s phone number. It may recommend books based on your past reading. It needs to know your location to list “toy stores near me.” The assistant requires access to your calendar to warn you of an upcoming event.

One technology trend that will partially ease privacy concerns is that the devices we own will have increasing computer power and memory over time. Many transactions and the data that drive them will be able to be executed on the device rather than in the cloud. We can expect some companies to emphasize privacy as a feature and take advantage of this trend in their devices. Beyond privacy advantages, the increased computational power within the local device will also mean that it has the additional benefit of handling many tasks without an Internet connection, making it usable for some tasks when a connection is not available.

Hacking of an individual’s digital assistant by criminals, law enforcement, or a government is an extreme privacy risk. Today, we see controversies such as government agencies looking at messages on journalist’s mobile phones to try to determine the source of leaks of classified information. This can be controlled to some extent by law or policy. But authoritarian countries are certain to use conversations with a digital assistant, as they do today with text messages, to determine if an individual takes anti-government positions.

Beyond privacy concerns, the control of content delivered by the assistant is a controversial issue. We see this issue today with companies like Facebook, Twitter, and web hosting companies trying to control content considered harmful, inaccurate, or deceitful. This becomes increasingly important as people become dependent on the assistant and as it becomes a part of childhood.

Authoritarian regimes that control content on the Web will certainly attempt to deliver propaganda through the assistants and silence any sources of information they don’t want to reach their population. This could be most harmful if it starts in childhood. The term “brain-washing” suggests the potential long-term effects of such control of an assistant that is almost a close friend and always there. This potential is likely to make authoritarian regimes promoters of selected digital assistants approved by the government.

Is the potential of evolving digital assistants for misuse outweighed by its benefits? This is a classical tradeoff in evolving technology. An obvious example is the concern by some experts that AI will take too many jobs or allow computers to eventually dominate humanity. Since the concern of technology eliminating more jobs than it creates has not historically proved accurate, we can hope that the downside of computer intelligence in general and digital assistants in particular is also overstated.

More fundamentally, concerns about improving technology have always succumbed to its advantages. We didn’t give up our cars when they caused smog; we addressed the problem with the additional technology of emissions control. Digital assistants will be adopted for their advantages. There will be a perception that someone who doesn’t have one is at a disadvantage to others. When we see a friend in a conversation ask the assistant in their mobile phone to settle a question that came up, it encourages us to use that capability. Do you stop using web search because the first selections that pop up are ads?

Technology has always changed what it means to be human to some extent. Automobiles have changed the meaning of distance. Phones have eliminated the need to be with someone to have a conversation.

But digital assistants connecting with us through language is a tighter connection with technology than some past innovations. As the technology improves and as humans get access to digital assistants at an ever-younger age, the assistants can augment human intelligence and provide us more direct control over other technology. To the degree that a human language interface eases the use of a new technology, digital assistants will accelerate adoption of such technologies, in many cases becoming indispensable to using the new technology.

Will digital assistants become part of being human? Do we have a choice or is that evolution inevitable?