For many, AI, whether it's in the form of a customer experience chatbot or helping autonomous car owners parallel park their vehicles, has eased some of the repetitive tasks of daily life.
For much of the public, AI has become an expected convenience, even though the term itself is somewhat vague. Software vendors routinely advertise product updates to provide more AI benefits without getting much more specific than that.
But amid the excitement and profits of a seemingly ever-accelerating industry, critics of AI, fueled by populist ideals and dystopian sci-fi movies, warn that AI, if left unchecked, could cause grave harm. Some AI skeptics fear that it could even destroy the world if AI-controlled military weapons get out of control.
As for AI developers and vendors, they have not scurried to make clearer what exactly AI can do, how it works and the amount of data it needs to create algorithmic models. A near total lack of regulation, coupled with users' willingness to give up data in exchange for convenience and productivity, has enabled AI vendors to operate with impunity.
Profits in the haze
In fact, companies like Amazon, Google and Facebook, while providing monetarily free, AI-powered tools and services to consumers, have not divulged much information about their technology and AI projects.
These tech giants have tried to conceal from local communities plans to build data centers in their areas and have misled customers about how their personal data is used. These practices, however infrequently they have been revealed, have only added to the confusion and lack of clear information about data and AI benefits and risks.
Yet, surrounded by uncertainty about AI benefits and risks, tech companies have reaped huge profits by developing and selling enterprise- and consumer-level AI tools and services. It's not a coincidence that the top ten highest valued corporations in the world include Microsoft, Apple, Google, Facebook and Amazon.
Those vendors will likely eventually be faced with a choice, said Luca De Ambroggi, senior research and analyst director at IHS Markit.
Luca De Ambroggisenior research and analyst director, HIS Markit
With the power of AI, people can work less and enjoy more leisure time or get paid more with AI automating so much work. Or, workers could lose their jobs or see stagnant wages as vendors continue to churn out new AI products and collect the profits that come with them.
AI systems "will be able to provide a lot of wealth," De Ambroggi said. "The wealth could be given to the hands of a couple of corporations or our society could be able to take advantage of this wealth and spread it to others."
It would be hard to make the case that tech companies haven't spread some wealth to society. Millions of people around the globe work in the tech industry, with millions more expected over the next few years. And they aren't all jobs for industry titans. As major technology corporations have continued to expand over the last several years, a new breed of business has sprung up at their feet -- the AI startup.
The connected startups
Thousands of startups, based all over the world, sell AI-driven tools and services aimed at saving time, cutting costs and boosting profits for businesses across all industries. Many are well-funded, raising millions after only a matter of months, and many provide unique and targeted services, filling the gaps that consumer and business-level tools from large organizations have sometimes been unable to fulfill.
AI-related companies raised $9.3 billion in 2018, a sharp increase from 2017, according to the "MoneyTree Report: Q4 2018" from PwC and CB Insights. The report noted that AI-related funding has increased every year since 2013.
Many AI startups can benefit from the tech giants, as many deploy their programs and services on big vendors' cloud platforms or use their development tools to build AI models.
Cogitai, for example, a California-based startup, offers Continua, an AI platform that can perform reinforcement learning on the cloud.
While reinforcement learning -- essentially a system of rewarding positive behaviors and penalizing negative ones that has been used to train unsupervised learning models -- is not a new concept, the company claims it's the first commercially available platform to use it.
"There's a lot going on in the academic space in this area, but we aren't finding a lot going on in the commercial space," said Mark Ring, CEO of Cogitai.
The startup spent years developing the platform, and, as of early 2019, it was in the process of acquiring customers.
"Our platform is not as data-dependent as supervised learning is," Ring said, meaning it can be used by clients across a variety of industries.
Yet, while the company developed its own technologies, Ring noted that it is "implementing everything on AWS." But, he claimed the company is "doing it in such a way that we can move away from AWS at any time," with the idea that it could be implemented on an organization's own system.
For startups like Cogitai, it can be difficult, and even economically foolish, to completely separate themselves from the cloud services of Amazon, Google, Microsoft and other big-name players, all of whom also provide many machine learning development and implementation tools on the cloud.
As Todd Lohr, principal at KPMG, put it, "When you're looking at the long-term AI players, it's going to be about the large cloud providers."
Outside of the cloud, tech giants are responsible for other fundamental AI development tools, as well. TensorFlow, a widely used framework for running machine learning and deep learning algorithms, was originally developed by Google before it became open sourced in 2015.
GitHub, a hosting environment used by organizations and independent developers alike to share and collaborate on code, was bought by Microsoft in 2018 for $7.5 billion.
There isn't evidence that these open source tools are being manipulated or abused by their corporate creators and owners, but it shows just how deeply integrated companies like Microsoft and Google are in the world of AI benefits.
What that will ultimately mean for the general public is unclear. Government rules and regulations around AI, if they ever become fully established, could one day make it necessary for these companies to break up some of their AI-related holdings. At the least, like the GDRP European law has attempted to do, regulations could help keep the public informed about how data and AI are being used and for what purpose.
For the average U.S. citizen, however, the more pressing issue might be the steady stream of jobs being handed over to AI-powered machines and software.
Potentially less work for more money
Indeed, the future of work will see employers increasingly "mapping tasks to skills" rather than people to jobs, possibly leading to a reduction of full-time positions, said J.P. Gownder, a Forrester analyst.
"If we have a very particular task that needs to be done, we can have it done by a full-time employee ... someone from the gig economy ... or we can have an intelligent machine, an AI, a bot or a similar entity take on that task," Gownder said in a keynote at IBM Think 2019 in San Francisco in February.
If the tech giants and enterprises "share the wealth," as De Ambroggi put it, that could mean people working part time or in the gig economy might get paid more for working less. They would make full-time salaries working what are now considered part-time hours.
Potentially, as AI eliminates busywork, "we could work five hours a day [and] enjoy life," De Ambroggi said.
That is, again, if corporations share more of the AI benefits. That remains to be seen.