Options to combat computer tech bias
Our July 8 entry (1) explained how Google and its YouTube subsidiary could slant the readily accessible internet content based on political considerations, and (2) discussed some of the accumulating evidence that they are doing so.
Now let’s factor in two more computer tech giants, Twitter and Facebook, and then discus what, if anything, should be done to limit the potential for abuse.
A. Completing the picture – Fueled by a lucrative share of internet advertising revenue, Alphabet Inc. (holding company for Google and other ventures) currently has a stock market value (market cap) of some $785 billion.
That doesn’t make Google the biggest computer tech company; Amazon, Apple Computer and Microsoft have even loftier market caps. But unlike Google, these companies are primarily involved in providing computer, etc. equipment and services to users (and in the case of Amazon shoppers) versus internet content hosting or search services.
Facebook and Twitter are major players in social media. Like Google, they have the power to favor or counter internet content based on political considerations. Here’s a quick rundown, including a sampling of controversies that have arisen.
#FACEBOOK – Users of Facebook (currently some 2.4 billion worldwide) share content intended for fellow members of loosely defined social communities. Once someone joins Facebook, they can contact others to seek acceptance as FB “friends” or vice versa. There are also procedures for “defriending” other people or blocking posts that aren’t welcomed.
Like Google, Facebook captures a lucrative share of internet advertising revenues. Its current market cap is $566 billion.
Facebook doesn’t create the content posted by users, but it is understood to have an obligation to filter out or flag objectionable content – as is being done in some cases. Some complain, however, that objectionable content is slipping through the net. Facebook, YouTube overrun with bogus cancer-treatment claims, Wall Street Journal, 7/2/19.
Weeks after getting out of jail in November 2017, [a man convicted of practicing medicine without a license] was back on Facebook. His main page gets frequent updates with posts selling his discredited cancer and dietary theories, plus services and products. Videos on an account featuring Mr. Young have earned more than 900,000 views, according to an analysis by social-media intelligence firm Storyful, which is owned by News Corp., the Journal’s parent company.
And here’s an interesting policy change, which could seemingly permit veiled death threats against people charged with certain crimes. So much for the presumption of innocence! Facebook updates standards to allow death threats against alleged sexual offenders, John Gage, Washington Examiner, 7/9/19.
There have been periodic complaints that Facebook has a liberal bias – and commitments by Facebook management to review their procedures, set up advisory boards, etc., which basically haven’t changed anything. Is Facebook too liberal? It pledges to investigate charges it's biased against conservatives, Jessica Guynn, usatoday.com, 5/3/18.
In September , President Trump boosters Lynnette Hardaway and her sister, Rochelle Richardson, known as Diamond and Silk, began noticing that Facebook was limiting the reach of their posts and preventing them from alerting followers to new videos on their page. After months of trying to wrangle an explanation out of Facebook, they say they received an email that their content was considered "unsafe to the community." *** [CEO Mark] Zuckerberg told lawmakers Facebook had made "an enforcement error," but that did little to tamp down the controversy.
Polling indicates that Facebook is the least trusted of the social media firms as a source of objective internet content. 76% of Conservatives Don’t Trust Facebook, Jim Hayek, americantruthtoday.com, 7/9/19.
More than three out of four conservatives say that they do not trust Facebook to treat all of its users equally, according to a new poll, which adds that more than half of respondents also said they hold the same views regarding Google, YouTube, and Twitter.
#TWITTER – Content providers log in to post “tweets” of up to 280 characters (which can be augmented by photos, links, etc.). These messages are theoretically addressed to all internet users. Anyone can receive them, retweet them and/or answer them by signing up as a follower, subject to the originator’s right to block unwelcome followers.
Worldwide, there are over 300 million active (at least monthly) Twitter posters and some 500 million “tweets” are posted per day. Twitter by the numbers, omnicoreagency.com, 1/6/19.
Twitter’s market cap is a comparatively modest $29B, which reflects their limited success in generating ad revenues from posting tweets. The company may be considering the rollout of a different business model to more effectively monetize its operations. Twitter is looking to diversify its revenue with a paid subscription, fool.com, Adam Levy, 3/23/17.
Here are some recent controversies concerning the use of Twitter, which in sum could be taken to suggest a bias against conservatives.
•President Trump blocked certain followers of his Twitter feed who had posted negative comments. Critics sued on grounds that the president’s Twitter feed is an official channel of communication so he can’t block followers for “personal” reasons, and to date their position has been upheld by the courts. Twitter blocking by Trump is unconstitutional, appeals court rules, Jessica Schneider and Katelyn Polantz, msn.com, 7/9/19.
In our view, this outcome makes sense since the president’s Twitter posts – as he has used them - doesn’t realistically represent an expression of his personal versus official views. Some observers have disagreed, however, on grounds that the decisions of a private sector businesses are not normally subject to this degree of constitutional scrutiny. “No shirts, no shoes, no service,” etc. Also, even if Trump appealed to the Supreme Court and won, critics could still access his tweets without being designated as followers and respond as they saw fit through their own Twitter accounts. Trump, Twitter and the First Amendment, Wall Street Journal, 7/11/19.
Twitter allowed Rep. Alexandria Ocasio-Cortez to block an anti-AOC group from following her tweets and posting negative comments. The critics promptly brought suit, seeking to apply the logic of the Trump ruling in reverse – it’s too early to speculate on the outcome. AOC Has Trump Twitter Block Ruling Turned Against Her, Lee Moran, yahoo.com, 7/10/19.
•Joe Biggs with 214,000 followers was banned from Twitter shortly after criticizing Antifa (a leftist group that has been involved in violent altercations in various locations around the United States) and making a comment that was susceptible to being interpreted as a call for violence. Conservative Commentator Joe Biggs Banned from Twitter after Questioning Antifa, Warner Todd Huston, godfatherpolitics.com, 7/6/19.
And the closest I ever got to expressing anything dangerous, was to say that conservative need to arm themselves for self-protection. But I said not to just go out and buy a gun, there’s more to it than that. You need to take classes, you need to understand, respect, and fear that weapon and know the law.
•As an add-on to its existing rules against threats of violence, Twitter is now banning the use of “dehumanizing language” to characterize members of religious groups. Similar added protections for other designated groups are reportedly under consideration. Twitter bans dehumanizing posts about religious groups, Hamza Shaban, Washington Post, 7/9/19.
Before it broadens the rule, however, [Twitter] said it needs to assess additional factors, including: how to protect conversations within marginalized groups in which people use “reclaimed” terminology and to ensure its enforcement takes context into account and reflects the severity of the violations.
B. Issue – For present purposes, let’s focus on identifying whatever action(s) might help to ensure Google search and YouTube are operated impartially. Bear in mind, however, that policy changes for these companies might also be considered for Facebook, Twitter, etc.
Discussion follows of three options: break Google up on grounds that it has gotten too big and effectively monopolized the market; mandate impartiality and audit compliance; or subject Google’s decisions to judicial (versus regulatory) review on a limited basis.
C. Antitrust – For critics who see the current dominance of Google and other computer tech companies as harmful and doubt that matters can be improved by government regulation, a forced break-up of Google (shades of the trust-busting activity that peaked in the early 1900s) may be seen as an attractive option.
Granted that the AT&T breakup in 1982 was the most recent example of reversing allegedly excessive business concentration by brute force, and that this supposedly beneficial transaction was partially reversed when some “baby Bells” started merging with each other. IBM subsequently fought off a proposed breakup, only to see its dominance in the mainframe computer world eroded by the rise of personal computers.
Nevertheless, there has been renewed interest in exploring the potential benefits of lowering the boom on Google et al. Breaking up one or more of these companies might appeal to antitrust agencies (Department of Justice, Federal Trade Commission), which are interested in justifying their existence, to some academics, and to politicians on both sides of the aisle. Tech Giants Google, Facebook and Amazon Intensify Antitrust Debate, Jacob M. Schlesinger et al., Wall Street Journal, 6/12/19.
The return of antitrust to the spotlight reflects disruptive political and economic forces. Many Democrats are moving to the left and looking more to government to address economic and social concerns. The Republican establishment is struggling to balance a coalition that includes both longtime business allies and President Trump’s populist base, which has at times declared open warfare on the U.S. Chamber of Commerce over outsourcing and immigration.
Last week, executives from Google, Facebook, Amazon and Apple testified before a House antitrust subcommittee, along with experts from academia and the legal community. The company witnesses reportedly came across as defensive and in some cases ill-informed. Congress grills Big Tech over competition, money and power, Barbara Ortutay & Matt O’Brien, middletownpress.com, 7/17/19.
The company representatives didn't help their case by pleading ignorance on multiple occasions. [For example,] Google's director of economic policy, Adam Cohen, said he was "not familiar" with how much Google pays Apple for the right to supply the default search engine for Safari on iPhones. (Rep. Jamie Raskin, a Democrat from Maryland, said It was $9 billion in 2018 and $12 billion in 2019.) Cohen also said he was "not familiar" with allegations of widespread fraudulent listings on Google Maps.
Similar hearings with a conservative thrust were conducted by a subcommittee chaired by Senator Ted Cruz (R-TX). Two previously discussed issues (Time to crack down on Google? 7/8/19) were covered, without convincing answers, including Google’s placement of scores of Prager University videos on the “restricted list,” and the abrupt dissolution of an advisory review panel for Google after some 2,500 employees protested the appointment of Kay Coles James of the Heritage Foundation. Rachel Guidice, daily signal.com, 7/17/19 & 7/17/19.
In our view, however, it would be hard to make a serious case for a Google break-up. Consider the following points:
•Although Google dominates the internet search business, there are other firms that can provide generally equivalent search results. Time to crack down on Google? 7/8/19.
•Google didn’t win the search business by engaging in anti-competitive behavior, it offered a superior search engine (which others have now apparently emulated). Why Yahoo lost and Google won, Gil Press, forbes.com, 7/26/16.
•Google’s service is “free.” Sure, the internet ads are annoying, but most people would probably rather put up with them than pay a subscription fee.
•The most serious concern about Google’s dominance is the evidence of political bias, which ultimately reflects the predominant political leanings of its workforce. Time to crack down on Google? 7/8/19.
And here’s some testimony from Dr. Robert Epstein (American Institute for Behavioral Research and Technology) that Google had a tangible effect on the 2016 presidential race – never mind its potential role in 2020 - even though Donald Trump won the election. Expert confirms Google swung upwards of 10.5 million votes to Hillary in 2016, John Salvatore, flagandcross.com, scroll down to video (1:57), 7/17/19.
It’s hard to see how this issue could be solved, however, by splitting the company into several pieces (initially staffed by subsets of the current Google workforce). So let’s consider some other options.
D. Regulation – The government could theoretically require changes in Google’s operations without the need for disruptive structural changes, e.g., take steps to ensure bias-free search results and screening of videos posted on YouTube. Sen. Josh Hawley (R-MO) has suggested a legislative solution along these lines.
Google and other computer tech firms currently enjoy legal immunity (under Section 230 of the Communications Decency Act of 1996) for facilitating internet access to content created and posted by others. What if Google and other major tech firms (let’s give smaller firms a break) were required to choose between (1) acting as a publisher, with potential legal liability for faulty content on their platforms, or (2) earning their immunity by undergoing periodic Federal Trade Commission audits to prove the political neutrality of their search algorithms and content restriction? GOP senator wants to remove key legal protection for tech giants, Allan Smith, chbc.com, 6/19/19.
Hmm, be a publisher or be an access facilitator, but you can’t have it both ways. The concept seems intuitively attractive, and if the FTC was charged with developing the regulations Congress wouldn’t have to worry too much about the details.
Section 230 currently provides, however, for Google et al. to play a dual role. Josh Hawley’s internet censorship bill is an unwise, unconstitutional mess, David French, conservativereview.com, 6/20/19.
• No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.
• No provider or user of an interactive computer service shall be held liable on account of any action voluntarily taken in good faith to restrict access to or availability of material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected.
Also, a requirement that Google et al. prove their operations are bias-free is impractical. Human beings have a facility for rationalizing their own thoughts and the outcomes in any non-random selective process will invariably be influenced by preconceptions and values, i.e., by biases. Mistakes were made (but not by me); Why we justify foolish beliefs, bad decisions, and hurtful acts, Carol Tavris & Elliot Aronson, 2007.
The basic model goes something like this: Mistakes start small and grow over time. When conflicting information or ideas surface, creating dissonance in the mind of the person(s) concerned, the natural human reaction is to minimize the dissonance (how could I, a wise and good person, be thinking or doing something wrong?) by justifying the ideas and/or actions that are in question.
What about the argument that Google’s search operations are run by algorithms so biases of Google employees don’t matter? Nonsense - search algorithms are created and adjusted based on human judgement, so their use by computers simply masks the bias issue rather than eliminating it. Time to crack down on Google? 7/8/19.
Accepting that current Google operations are inherently biased, what could the institution of government audits accomplish except to substitute one set of biases for another? That wouldn’t necessarily represent a step in the right direction. Josh Hawley’s internet censorship bill, op. cit.
It’s often the case in Washington that the title of a bill communicates the exact opposite of its content or effect. [Thus, Sen. Hawley’s Ending Support for Internet Censorship Act] would grant [the federal government] enormous power over social-media content. It would enable public censorship in the name of limiting private control.
OK, the Hawley bill isn’t the answer either. Any other ideas?
E. Judicial review – As already discussed, Section 230 of the Communications Decency Act exempts the computer tech companies from liability for the content of communications that they did not create. The Hawley bill would limit this privilege by establishing oversight by the Federal Trade Commission, but a case could be made for creating a role for the federal courts instead.
Say the rule that computer tech companies aren’t responsible for content they did not create was continued, but the license to restrict access to or availability of material – now subject only to a requirement of good faith – was additionally conditioned on not interfering with the rights of free speech, etc.
This approach would not magically resolve all disagreements about the conduct of the computer tech giants, there isn’t any way to do that, but it would make the federal courts (appointed judges, several levels of review) responsible for ruling on disputes instead of a single bureaucratic organization. It would also avoid the mistake of making the computer tech giants targets for massive monetary damage claims.
Also note that the courts are getting involved in this area anyway, as shown by rejection of the president’s attempt to block unwanted “followers," so it may make sense to give them a defined role instead of allowing them to free-lance.
Do you readers agree, or should other approaches be considered?