Congress hasn’t determined the artificial intelligence behavior it wants to regulate

Congress’ attempt at AI legislation

The 116th Congress has three bills or resolutions of note sitting in a number of committees that attempt to address the use of artificial intelligence in American society.  H.Res. 153 is intended to support the development of guidelines for the ethical development of artificial intelligence.

Meanwhile,  HR 2202 seeks to establish a coordinated Federal initiative to accelerate artificial intelligence research and development of the economic and national security of the United States, while H.R. 827 is concerned about training and retraining workers facing employment disruption by artificial intelligence.

The bills and resolution are more exploratory versus regulatory in nature.  They don’t expressly get at any behavior that needs to be regulated.  The nascent characteristic of artificial intelligence development along with government’s inability to keep regulation at the same pace as technology development may be in part the reason for the “let’s see what we have here” stance of early regulation.  In other words, Congress may just be getting a feel for what AI is while taking care not to interfere with the innovation needed to further develop the technology.

On the other hand, these bills could help Congress get ahead of the issue of labor disruption. Workers read about AI’s capacity to replace jobs that are more pedestrian or mundane; jobs currently occupied by lower income workers.  The prediction about labor disruption won’t leave certain higher-waged jobs untouched either as AI platforms, data mining, and machine learning threaten professions such as accounting or law.  By determining the data needed to thoroughly analyze the impact and growth of artificial intelligence; identifying the industries that will benefit most from or be harmed the most by AI; and comparing today’s existing job skills with the job skills that will be needed in order to work alongside AI, Congress contribute to alleviating labor force disruption.

The Human-Machine Relationship

Unlike a number of dire predictions of the emergence of “SkyNet” and terminator-like machines subjecting humans to slavery or worse, most analysts and commentators see AI as a tool that augments human capabilities, making humans better or more productive.  The emphasis will be on collaborative intelligence, with human and machine working together.  And how well that relationship works depends on how well humans program the machine.

Another consideration is artificial intelligence’s ability to self improve.  The goal of AI development is to build an artificial intelligent agent that can build another AI agent, preferably with even greater capabilities.  The vision is to move from AI’s narrow, single-task capabilities to a more general AI, a concept that sees AI exhibiting greater abilities across multiple tasks, much like humans.

Possible targets of legislation or regulation

If legislation or regulation is to target the machine-human relationship, elected officials and regulatory heads will have to consider their policy initiatives impact on:

  • the ability of humans to train machines;
  • the ability of humans to explain AI outcomes;
  • the ability of machines to amplify human cognitive skills;
  • the ability of machines to embody the the human skills necessary that lead to the extension of human physical capabilities;
  • the ability of artificial intelligence to self-improve;
  • the difference between “safe” AI (the ability to maintain consistency in AI outcomes) versus “ethical” AI (ensuring that AI platforms are not a threat to human life.

Conclusion

Just like the application of artificial intelligence, Congress’ foray into regulation of AI is nascent.  This is the time for AI’s stakeholders to either begin or maintain efforts to influence all levels of government on AI’s use in the commercial sector.

 

Advertisements

How to engage the regulator

About fifteen years ago I worked for a local government agency that was responsible for regulating basic cable rates and managing the franchise agreement the county had with two cable providers.  One day I received a call from a Bear Sterns analyst. He wanted some information on the probability that the county would approve a cable company’s new franchise, if I recall correctly.  While we did not get too many calls from investment bank analysts, I was not completely surprised that an investment bank would take interest in a regulatory matter, especially one that would impact the cable company’s revenue.

What does surprise me, however, is how little the investment community appears to know about how people in government think.  This lack of knowledge may start as far back as business school.  Take a cursory review of an MBA program’s curriculum and there is a dearth of coursework on how government operates or the impact of policy making on private sector revenues.

I don’t know the specific reason for this gap in knowledge.  I can hazard the guess that the investment sector like the general public sees government as that inconvenience, that necessary institutional evil comprised of soulless rules and laws and unaware bureaucrats that slows down innovation while imposing undue tax burdens.  The cynic in me would agree with this overall assessment, but I would caution that any member of the investment community engaging a regulator stop short at viewing the staff member or ultimate decision maker in a government agency as “soulless” or “inconvenient.”  That approach won’t get you far, especially when trying to either anticipate an event, understand a policy, or influence an agency’s proposed regulatory action.

Government and politics is a people business. With all the talk of artificial intelligence and machine learning, government decision making is a human endeavor, with agency heads and staffs taking into account the positions of various stakeholders, all of whom are human as well.  While being “best friends” with a government analyst is not necessary to effectively extract information or exercise influence, understanding an agency’s mission and its decision making methodology or mechanism is important.  Knowledge of a staffer or decision maker’s educational background can also help. Being a fellow alum can be an icebreaker.  Also, understanding past decisions and the rationale behind them can help in anticipating future outcomes.

Also, keep in mind that the agency staffer or decision maker is interested in the private sector as well.  The staffer may have an interest in the evolution of the product she regulates. For example, while a staffer at the Florida Public Service Commission I found the emergence of the internet fascinating. Watching telephone companies and cable companies position themselves and develop the technology that would later power what is today called broadband was a great experience.

If that interest on the part of a staffer is recognized by a private sector player, an opportunity is created to make more robust, detailed presentations because an interested staffer is one that has done her homework about the service being regulated.

Bottom line, engage the regulator as more than someone presenting a legal or regulatory barrier or espousing rules.

Capital. The true digital divide

A couple early morning thoughts on the digital divide.  So far the digital divide narrative has occupied two schools of thought that are not necessarily opposed to each other.

Race and the Digital Divide

The first school of thought revolves around race.  Given that within the black American community there is a higher level of poor households, affordability is keeping blacks from accessing the internet via high-speed broadband infrastructure.  If blacks do not have the income to sustain a broadband business model, then internet access providers are less likely to deploy facilities in poor neighborhoods.  Lack of deployment in these neighborhoods may result in a barrier to valuable information that may lead to greater economic opportunities, according to advocates seeking to close this gap.

Rural Communities and the Digital Divide

The second school of thought revolves around rural communities.  The argument is that lower population density as compared to urban areas makes deploying broadband access facilities in rural areas more expensive.  In addition, terrain, such as that faced by internet access providers in mountain states, has traditionally added to the problem of higher costs to provide broadband access facilities.

An Overlooked Divide

There is another divide, one that is often overlooked and it has to go to what is known as “first-mover advantage.” The real value generated by the internet is the ability to extract, analyze, package, and distribute information, and have that information be available digitally forever.  The focus on a gap between facilities deployed in black neighborhoods versus facilities deployed in white neighborhoods or the gap between rural community deployment versus urban community deployment goes to seeking out new suppliers of information.  The civil right veneer that has been placed over the broadband racial divide hides this supply-side characteristic from the policy debate.  It has also created the opportunity for the political left to craft an electoral package that can be sold to voters.

It is the other side of the equation, the production side, that, in my opinion holds more value.  When we look at the history of the internet, particularly the period when the internet was commercialized, its players included white venture capitalists; Web 1.0 internet service providers, i.e. AOL, CompuServ, Mindspring, etc.; and dial-up access providers such as BellSouth.

Black Americans could always access information from analog sources, i.e. television; print media; or word of mouth, but the efficient extraction, cataloging,  indexing, aggregation, and distribution of information via the internet were the domain of companies invested in and managed by whites.  As whites continued to level their first-mover advantage, this gap between producer/owner of capital and consumer continued to grow.

Capital not only seeks a vacuum, it also seeks a return.  Returns from investing in black or even rural communities were not going to be as high as returns invested in affluent neighborhoods, neighborhoods whose residents probably owned shares in the very companies that commercialized the internet in the first place.  Closing the “digital divide” means first closing the capital divide.

What will Government Do Next?

Government will do nothing from a capital perspective to close the digital divide. The Federal Communications Commission has a number of universal service funding initiatives designed to encourage mobile and fixed broadband deployment in rural areas; to facilitate the delivery of health care via broadband; and to reduce the costs incurred by low-income consumers for accessing and maintaining high-speed broadband service.  By subsidizing the consumer demand for broadband services, the Commission hopes to encourage the delivery of broadband services.  But again, the focus is on consumer demand, not bridging the capital gap.

The philosophical underpinnings of the American economy, where capital is to flow freely to its best use may prohibit government from taking any concrete action for closing a capital gap.  If blacks or rural residents had sufficient capital to purchase, construct, or maintain broadband access facilities, using their intimate knowledge of their communities to distribute services, we might see a decrease in the gap.  We should expect that government will stay on a path of incentivizing capital investment in infrastructure development versus trying to repair capital discrepancies via a capital transfer.

Should Comcast and Verizon be allowed to enter the information mining game? Yes.

Overview

Proponents of the implementation by the Federal Communications Commission of net neutrality rules have been expressing outwardly that by ensuring no throttling of traffic from websites; no blocking access to favored and lawful websites; transparency when it comes to terms and conditions of service or network management; and the prohibition of favored treatment of one content provider’s traffic over another, that consumers of broadband services will be treated fairly and that edge providers will be able to innovate on the edge of the internet while competing with core providers.

While proponents have successfully convinced millions of Americans that net neutrality is about the consumer’s ability to democratize the web or have their voices equally heard among other, especially larger corporate voices, the real issue is whether core providers should be allowed to participate in the information markets or be kept out by making an 85-year old statute a barrier to their entry.

Battle in the Information Market

The statutory approach recommended by edge providers such as Facebook and Google to ensure that core providers such as Comcast and Verizon are reigned in is to apply Title II of the Communications Act of 1934. Edge providers make their bread and butter by mining information from visitors to or users of their website services and packaging that information into advertising products that they sell to businesses that are trying to get their services before as many eyeballs on the internet as possible.

The concern that the edge provider has with the core provider is that given the core provider’s “gateway” service and the core provider’s alleged monopoly or near monopoly control of the access to the internet, the core provider will then be able to capture consumer behavioral information that the edge provider has less access to.

The core provider, the edge provider will argue, is gathering this information from its telecommunications infrastructure; therefore, to ensure fairness, the core provider should not be allowed to call the telecommunications portion of his service an information service just so that they can skirt the information or data collection requirements under Title II.

By creating a net neutrality rule that says that core providers should treat access as a telecommunications service, the edge provider gets the government to apply a barrier to entry to the information market, a barrier that the edge provider has no confidence its superior information services can erect itself.

The Content Endgame: What Would Title II Do and Not Do

If Verizon wanted to use information “that relates to the quantity, technical configuration, type, destination, location, and amount of telecommunications service used by a consumer of a telecommunications service, that information, in general, would be limited in use to services related to the provision of telecommunications services. Verizon would not be able to use information related to the provision of telecommunications services to predict consumer demand for Verizon’s video streaming services.

Interpreting and applying Title II in this manner would help Hulu and Netflix keep Comcast and Verizon at bay. What it may also do is expose Hulu and Netflix’s pricing and cost structures during any public hearing resulting from Hulu and Netflix’s new roles as consumers of telecommunications services. Sections 204 and 205 of Title II provide the Federal Communications Commission the authority to set just and reasonable charges and to have hearings on those charges or on complaints regarding charges and price schedules. What Hulu and Netflix may not understand is that regulation of a market means scrutiny of both of its sides, and challenges to rates charged by a core provider means rebuttal that could include discovery of what economic rationale underlies an edge provider’s assertions. In short, Title II opens the Pandora’s Box for edge providers, too.

What Title II doesn’t do is tell Comcast or Verizon that they couldn’t collect consumer behavioral information from their websites or information portals. This “oversight” is further evidence of how arcane Title II is. A declaration by the courts that a core provider’s services are information services, from end user through a core provider’s entire network would be indication that the State recognizes that core and edge providers equally play in the information markets. Avoiding a balkanized, bifurcated view of broadband service provision would make regulation of advanced communications more efficient because of less time spent having to look at two separated pieces of internet service versus one.

The FCC’s Constitutional Quandry

But even if regulators continued down the two-prong path of regulating core providers, the end game would still be how to treat the content portions of their services. The Federal Communications Commission should not want to be in the position where it would take a hands-off approach to Facebook’s information mining techniques while taking a heavy handed approach to Verizon’s emerging content play. It would cause a constitutional dust up were the Commission to regulate the content of one service provider but not the content of another.

Conclusion

Core providers have the technical knowledge and desire to enter information markets and for that reason alone they should be allowed to profit from the development of content and the extraction and packaging of data that drives a modern economy. Core providers shouldn’t be punished because their basic business model includes the conveyor belts that information is placed on when being extracted. Imagine telling a coal miner that in order to foster competition, they will have to forego their conveyor belt and, like a firm that entered the market late and poorly capitalized, will have to use their hands and wheel barrel to move coal out of the mine. That is not competition but favoritism.

The “economy” is doing better but I am seeing more homeless in Atlanta

I am seeing more homeless people in my West End Atlanta neighborhood. I have seen at least one sleeping in his vehicle. Others make use of the parks to sleep at night.  What I see on the ground does not coincide with the claims made in Washington of a booming economy.

WABE, citing data collected from the city of Atlanta, reported that the homeless population numbers around 3,000 people and is allegedly on a decline.  And last year, the Atlanta Journal Constitution reported that Atlanta ranks among America’s neediest cities based on 21 metrics including child poverty and the number of uninsured. Homelessness is the result of a number of factors including the lack of affordable housing, poverty, discrimination, and shifts in the economy. Can city policies adequately impact these factors?

Take the factor of affordable housing. Atlanta mayor Keisha Lance Bottoms has made affordable housing one of her top public policies, but it appears to me that such an approach falls out of line with one important goal of a city: to generate tax revenue necessary for providing the amenities that keep citizens interested in living in Atlanta.  Land owners want to see property values rise and see an increase in the revenues that their properties generate.

Also, as city leaders continue their efforts to make Atlanta a job center, they have to keep in mind that as part of the efficiencies offered by a city is the location of housing close to job centers.  Housing located close to job centers may also end up being some of the most costliest housing.

I ride into Buckhead every day from southwest Atlanta. I have blogged before about how the MARTA train feels more like those conveyor belts loaded with coal that go into a furnace to fuel a production facility.  In this case the human coal are the lower and middle income individuals heading into Buckhead to work a job that, ironically, may be on the chopping block in a few years due to artificial intelligence.  If these people can’t afford to live close to an employment center where they can walk to work, the pressures of living will really increase when they have to find alternative employment.

But even with current employment, there may not be enough affordable housing available because landlords will be under pressure to meet rising property taxes resulting from the increased values of their properties, at least in the short run. This rise in value and ensuing property taxes will result from increased demand for housing that Atlanta expects to face over the next ten years.

Let’s not forget the upward pressure expected on interest rates over the next two years.  Property owners will have to increase rents in order to cover higher mortgage rates.  For the city of Atlanta it means higher bond servicing costs as the city continues to raise money through bond issues for its development and operational needs.

Affordable housing, because of the above pressures, won’t increase in supply.  Only an economic downturn may bring about cheaper rentals but even that will be short lived because a downturn in the economy means a slowdown in hiring and the specter of non-affordability due to increased lost income.

Politics wise, it is time for elected officials, particularly Democrats, to eliminate the affordable housing mantra from their campaign slogans.  They won’t be able to achieve it at any meaningful scale.

 

Trump and the Federal Reserve: Governing with transparent purpose

Listening to the rhetoric of President Donald Trump over the past 19 months, if I were to summarize the role of government, it is to defend national borders, sustain an environment that creates jobs, and be impactful in driving up stock market values.  Mr. Trump has effectively drowned out the Republican congressional leadership to the point where I don’t care what Senate Majority Leader Mitch McConnell or Speaker of the House Paul Ryan’s views on what the government’s role is supposed to be.

Under my interpretation of public administration, the buck, when it comes to governing, begins and ends with who is in the White House.  It is the Executive who enforces the law and interprets the law every day given a particular problem.  An argument can be made that during the run-up to the 2020 presidential election, the views of Mr. Trump’s challengers will take on some importance as voters compare the record of Mr. Trump with the promises of his Democratic challenger, but Americans have a way to go before the Democrats settle down on a few contenders and beginning pushing their messages before the electorate.  All we have right now are the whispered names of Andrew Cuomo, Elizabeth Warren, Joe Biden, and, yes, Hillary Clinton.

I suspect that none of the above named Democrats will be serious contenders in the spring of 2020 anyway.  Listening to the roll call of potential presidential candidates is like believing that the baseball team leading their division eight weeks into the season will be in the World Series much less holding the trophy.

In my lifetime, Mr. Trump has been the most transparent of presidents when it comes to the factions that he promotes.  Mr. Trump has been consistent and clear with his America’s economy first message. He took Mr. Trudeau out to the woodshed during the renegotiations of the North American Free Trade Agreement.  He kept his word on pulling the United States out of the Paris accords on climate change and the Trans Pacific Partnership Agreement.  He will not enforce the mandate that taxpayers are required to purchase health insurance, facing penalties if they don’t.  These initiatives are driven by a philosophy of American economic nationalism with the hopes of creating incentives for American businesses to repatriate jobs and cash to America’s shores.

He’s recently been transparent about the most important engine in the American economy: the Federal Reserve.  Mr. Trump disapproves of the Federal Reserve’s increase in the target for its federal funds rate, even though the Federal Reserve’s independence gives the central bank the okay to thumb their noses at the President.  The federal funds rate is the interest rate at which the Federal Reserve’s member banks may lend each other money overnight.  Changes in the fed funds rate seep into the overall economy in the form of mortgage rates, credit card rates, and interest rates on bonds.  Higher rates raise the costs of borrowing making it tougher for businesses to invest in growth including hiring more labor.

Higher rates mean that the economy’s “labor to tax conversion mechanism” becomes less efficient.  The labor to tax conversion mechanism is that layer of the economy where companies convert human resources into tax dollars by adding labor to payroll and collect and transmit income, payroll, and social security taxes to the Treasury.  Tax dollars are collected by the U.S. Treasury and either deposited for future spending on public programs or to service the debt.

But as I alluded to before, companies will feel constrained by interest hikes as they see revenues and profits reduced by higher costs for doing business. This may mean, depending on the business, a move toward automation in order to reduce labor costs.  Taking labor off of payroll means removing a head that could be taxed.  Will government have to apply some type of alternative tax applicable to an artificial intelligence that replaces a human intelligence on a factory line?

Going back to transparency, neither Mr. Trump or any leading Democrats have clearly demonstrated an ability to describe to the American public how their current economic environment works.  Neither begin any of their discussions on the economy with a discussion on capital or describe how the central bank is still the only game in town and the relationship to and importance of the central bank to all Americans. Mr. Trump has come the closest which means at this time he is the only elected official that gets it.

 

Does Facebook’s business model disrupt the political information markets?

Facebook is engaging in a war against misinformation and divisiveness in the United States as perpetrated via social media, according to published reports by Bloomberg and The Atlanta Journal Constitution. Having done a 180 degree turn from its position last year that its platform was not used to cause a disruption of public opinion leading up to the 2016 presidential election, Facebook is using artificial intelligence tools to identify inauthentic posts and user behavior.  With teams comprised of data scientists, policy experts, and engineers, Facebook is blocking fake accounts and vetting news stories posted on its site.

Critics doubt that Facebook’s attempts to thwart future social media influence will outweigh its incentives to distribute fictional political stories that keep people glued to Facebook while providing advertisers with millions of pairs of eyeballs.  Facebook, according its 10-K annual report, garners almost of its revenues from advertising.  In 2017, advertising made up 98% of Facebook’s revenues.  According to Facebook’s 10-K, at the top of the list of factors that could adversely impact advertising revenues: decreases in user engagement, including a decline in the time spent using the company’s products.

Having used Facebook for eleven years, I witnessed the increase in the use of the platform as a tool for political engagement.  Facebook has expanded opportunities for voters to vet politicians and their policies.  I have seen a significant number of posts, including memes and video, that got the facts wrong; that showed no knowledge of process, politics, or economics.  Cynicism, fear, passion, inaccuracies, sincerity, patriotism, anarchy, and indifference all run rampant on Facebook.  But do I buy the argument that messages placed on Facebook by Russian agents spread so much misinformation that America became suddenly divided overnight? That “Russian interference led to a Trump victory?

No.  The divisiveness was already there.  Giving a couple hundred million Americans the ability to quickly share their thoughts, accurate or not, on the political news of day simply tore away the scab.

Further evidence of divisiveness in American politics: print, broadcast, and cable media.  American media is meeting the demand of a divided public, with Fox News occupying the Right and MSNBC and CNN serving the frenzied Left.

What Washington may truly be afraid of is that politicians have less control over the channels through which they are vetted.  On the one hand, Jeffrey Rosen, president of the Constitution Center, shared the following with The Atlantic’s Jeffrey Goldberg:

“Twitter, Facebook, and other platforms have accelerated public discourse to warp speed, creating virtual versions of the mob.  Inflammatory posts based on passion travel farther and than arguments based on reason.  We are living, in short, in a Madisonian nightmare.”

On the other hand, Americans may be taking to Facebook, YouTube and Twitter in search of alternative opportunities to criticize the political packages and action plans that politicians offer in exchange for votes and increases in taxes.  The divisiveness may be stemming from an increased lack of enchantment with democracy itself.  After all, according to Professor Yuval Harari, democracies are “blips in history” depending on “unique technological conditions” and losing credibility as democracy faces more questions about its inability to provide for and maintain a middle class.

Democracy is hard up to explain why almost all the nine million jobs created post recovery from the 2007-2009 recession have been “gig work” paying little to no benefits.  Democracy has yet to come up with a solution to a wealth gap that the Left invests time in describing, laying blame at the feet of the rich yet coming up with no solutions for a society that prides itself on equal access to the ballot but still comes up short on adequate access to capital.

To the question whether Facebook’s business model has disrupted the political information markets, I would, for now, answer yes.  Facebook has contributed to bringing unreasonable, uninformed voices into the arena. I for one do not want to be lead or have policy fed by impassioned, unreasonable voices, no matter what part of the spectrum they fall on.  What the political class may have to look at for in the near term is that democracy may be less of a facilitator of a peaceful transfer of power between its factions as the mob continues to peel away the scab.

 

 

Government’s role in regulating access to personal data

Yuval Noah Harari recent wrote an article for The Atlantic where he posed the question, “How do you regulate the ownership of data?” Professor Harari argues in the article that data is the most important asset today, moving ahead of land and machinery.  “Politics will be a struggle to control the data’s flow”, says Professor Harari.

Last spring saw the United States Congress’ struggle to at least map out a course through the turbulent waters of data privacy as members of the House of Representatives and Senate took the opportunity to grill Facebook CEO Mark Zuckerberg about his company’s handling of personal data obtained from the social media giant by a consultancy.

Part of this struggle may be due in part to the popularity of social media network platforms. Facebook has climbed from a digital bulletin board developed in the early 2000s in an Ivy League college dorm room to a global subscribership of over two billion people.  Former president Barack Obama’s Twitter following is in the millions while the current president, Donald Trump, is not shy or slow to taking to Twitter to either connect with and inform his base of supporters or attack the traditional media for what he perceives as unfair coverage of his administration.

Professor Harari notes that users of social media network platforms have not reached the point where they are ready to stop feeding the “attention merchants.”  Speaking on the difficulty subscribers may have in exchanging personal data for “free” services, Professor Harari points out that:

“But it, later on, ordinary people decide to block the flow of data, they are likely to have trouble doing so, especially as they may have come to rely on the network to help them make decisions, and even for their health and physical survival.”

Professor Harari offered up one solution, nationalization of data, to stem the abuses that corporations may impart on addicted social media and internet consumers, but admits that just because an asset is in the hands of government doesn’t mean things will necessarily go well.  Hence the question, how should the ownership of data be regulated?

The question will require public policymakers and politicians go through the exercise of defining “personal data.”  Would personal data be any characteristic about you? Would it be about any marker, no matter how temporary or permanent, that can be attached to you?  Must the “data” be something that the consumer actually produced?

Politically, attention merchants would want a narrow reading of the definition of personal data.  A narrower reading of personal data means being able to obtain more information pursuant to fewer restrictions. While this outcome would be ideal for corporate entities in the business of brokering data, I don’t see Republicans, even with their mantra of promoting business, enthusiastically endorsing less restrictive collection of personal data given the public’s concern for privacy.