artificial intelligence

Subterms

    More stories

    • 100 Shares109 Views

      in Elections

      Chaos and Concessions as Kevin McCarthy Becomes Speaker

      by

      More from our inbox:Should Babies Sit in First Class on the Plane?A Chatbot as a Writing ToolSupport Family Farms Haiyun Jiang/The New York TimesTo the Editor:Re “McCarthy Wins Speakership on 15th Vote After Concessions to Hard Right” (nytimes.com, Jan. 7):So Kevin McCarthy is finally speaker of the House. It took 15 votes to get him there.But considering the concessions he had to make, the unruly nature of right-wing Republicans and the razor-thin margin, the next two years are likely to be a nightmare for Mr. McCarthy.Sometimes be careful what you wish for.Allan GoldfarbNew YorkTo the Editor:It’s easy to blame Republicans for the debacle of the House leadership vote and all its predictable miserable consequences. But where were the 212 Democrats in all this?Sure, I can see the rationale behind a show of support and unity for Hakeem Jeffries at the outset. He’s much deserving and would have done a fine job. But that’s a battle Democrats were never going to win.Deep into the voting rounds when it became apparent that there would be no win for Kevin McCarthy without further empowering the right-wing extremists, wouldn’t it have been smarter for Democrats to have gotten together to nominate some (any!) moderate Republican and hope to deny both Mr. McCarthy and the extremists their day?Democrats are just as bad as Republicans in putting party loyalty ahead of what’s best for the American people.Russell RoyManchester, N.H.To the Editor:Re “How a Battle for Control Set the Table for Disarray” (news article, Jan. 8):As Emily Cochrane points out, in getting elected speaker, Kevin McCarthy accepted making changes to the rules of the House that are not merely a weakening of the powers of the speakership, but also a danger to the country. If Congress cannot agree to raise the debt ceiling, the United States could default on its debt for the first time. The mere threat is a clear and present danger.Is it possible that some Republican members of the House could, even though they voted for Mr. McCarthy, nonetheless join Democrats in voting against the most dangerous changes in the rules?If, instead, all House Republicans regard their vote for Mr. McCarthy as a vote for the concessions he made to become speaker, then each and every one of them has as much responsibility for the damage these rules will do the country as the radicals who insisted on the changes.Jeff LangChapel Hill, N.C.The writer is a former chief international trade counsel for the Senate’s Committee on Finance.To the Editor:What a day. I imagine that the second anniversary of the Jan. 6 insurrection will go down in history as the day the Democrats commemorated all the patriotic heroes who fought to save our democracy, while simultaneously the Republicans in Congress could be seen doing their level best to destroy it.Sharon AustryFort WorthTo the Editor:It’s not just the far-right representatives who can disrupt the workings of the House. The concession to change the rules to allow a single lawmaker to force a snap vote to oust the speaker gives the Democrats a filibuster-like power.If they want to stop a particular vote for a Republican-sponsored bill, all the Democrats have to do is keep calling for votes to remove the speaker. That vote would take precedence until the Republicans give up and take their bill off the agenda.By insisting on having the power to disrupt the workings of the House, the far-right Republicans have given the same power to the Democrats.Henry FarkasPikesville, Md.To the Editor:Teachers seeking to explain to their students the meaning of a Pyrrhic victory, look no further than Kevin McCarthy!Peter RogatzPort Washington, N.Y.Should Babies Sit in First Class on the Plane? Brian BritiganTo the Editor:Re “Um, Perhaps Your Baby Will Fit in the Overhead Bin?” (Travel, Jan. 7):This article has particular relevance for me, as someone who has traveled more than 100,000 miles every year for the last 25 years. I have seen a number of variations on this theme of babies in first class.The alternative to having one first-class or business seat with an infant on one’s lap is to buy two seats or even three seats in coach, which allow for the parent to have the option of holding the child or placing the child in a travel seat. It would also be fairer for airlines to require that parents buy an actual seat for an infant when it comes to purchasing seats in business or first class.There is a clear difference between a domestic first-class cabin for a two-hour flight and an overnight transcontinental flight where the entire point of paying $5,000 for a seat is to be able to sleep so one may function the next day during back-to-back meetings.My heartfelt advice to those parents contemplating their options is to buy a Comfort Plus or premium coach seat for you as well as for your infant to have ample space and to be a good citizen.Ronnie HawkinsWashingtonTo the Editor:A few years ago, my husband and I flew on Scandinavian Airlines from D.C. to Copenhagen. There were perhaps half a dozen babies on the plane, but we heard not a peep from any of them for the length of the flight. Why? The plane had fold-down bassinets in the bulkheads, and people traveling with babies were assigned those seats.Of course, there are no surefire ways to prevent disruptive passengers, whether they’re children or adults, but the airlines in this country disregard their own role in this mess by making flying such a miserable experience for everyone.Debra DeanMiamiA Chatbot as a Writing Tool Larry Buchanan/The New York TimesTo the Editor:Re “Fourth Grader or Chatbot?” (The Upshot, Jan. 4):I have been a teacher of writing for the past 38 years, and my first reaction to ChatGPT, a new artificial intelligence chatbot, was dread: How could I prevent my students from using this technology? My second reaction was to wonder how I might use it myself.Once we are done with denial and hand-wringing, teachers need to think about how we can use A.I. to help teach student writing. This tool can help generate ideas, offer suggestions, map out structures, transform outlines to drafts and much more that could demystify the writing process.The technology is here to stay; our job will be to advance the education of our students by using A.I. to develop their writing and thinking skills.Huntington LymanMiddleburg, Va.The writer is the academic dean at The Hill School in Middleburg.Support Family Farms Antoine CosséTo the Editor:Re “What Growing Up on a Farm Taught Me About Humility,” by Sarah Smarsh (Opinion guest essay, Dec. 25):I am just one generation removed from the family dairy farm, and my cousins still operate one in Idaho and their lives are tough. In the words of Ms. Smarsh, they’re “doing hard, undervalued work.”Ms. Smarsh makes a strong case against giant agricultural corporations and their “torturous treatment of animals.”Currently, the majority of farm production is driven by corporate greed. However, small-farm, organic-raised meat and produce are expensive alternatives, which are out of reach for low-income, food-insecure families.More moral, sustainable food production is a policy issue that our lawmakers should address. Ms. Smarsh is right: Family farms are being “forced out of business by policies that favor large industrial operations.”Mary PoundAlexandria, Va. More

    • 113 Shares199 Views

      in Elections

      Should Merrick Garland Reveal More About the Mar-a-Lago Search?

      by

      More from our inbox:Democrats’ TacticsThe Robot TherapistFamily PlanningFormer President Donald J. Trump could oppose the motion to release the warrant and inventory of items taken from his home, and some of his aides were said to be leaning toward doing so.Emil Lippe for The New York TimesTo the Editor:Re “Attorney General Stays Quiet, as Critics Raise the Volume” (news article, Aug. 10):The Justice Department really needs to explain to the American people why the F.B.I. searched former President Donald Trump’s home, given the precedent-shattering nature of what happened. It should do so for three reasons.First, given that such an act has never occurred before in American history, the public deserves to know why a former president was sufficiently suspect that the F.B.I. felt it had no choice but to conduct a search of his living quarters.Second, the silence will be interpreted and misinterpreted on the basis of partisan biases. Already right-wing leaders have deemed this an act of war, while liberals perceive it as justified, given the president’s predilection to illegally hold onto classified materials. To correct misperceptions, the D.O.J. needs to explain its rationale.Third, there is precedent for this. In 2016, James Comey, then the F.B.I. director, sent a letter to Congress to explain why the bureau was investigating Anthony Weiner’s email messages, which bore on Hillary Clinton’s presidential campaign.If a Justice Department official went public in a case like that, surely it should offer an explanation for a case this precedent-breaking and important.Richard M. PerloffClevelandThe writer is a professor of communication and political science at Cleveland State University.To the Editor:Like many other Americans, I’m curious to know more about the Justice Department’s investigation of Donald Trump. But I think Attorney General Merrick Garland is right to keep silent about the details at this point. Mr. Khardori cites “exceptions” to the prosecutorial rule about not commenting on ongoing investigations, but none of them apply particularly well here.We already know what it’s appropriate for us to know at this point, such as that the search of Mar-a-Lago had to have happened only after a federal judge agreed that evidence of a serious crime was likely to be found there.In due time, I suspect, we’ll know a lot more. For now, let’s be patient and let the Justice Department do its job. The list of reasons for it to avoid public comment at this stage is longer than the list of reasons for it to do the opposite.Jeff BurgerRidgewood, N.J.To the Editor:“He Wielded a Sword. Now He Claims a Shield” (news analysis, front page, Aug. 11) certainly gets it right when it notes that the current outrage of the former president and his supporters over the F.B.I.’s execution of a search warrant at his Mar-a-Lago estate brings up echoes of his past behavior.After all, for Donald Trump, if he loses an election, someone else rigged it.If the U.S. Capitol is attacked, someone else incited it.Taking the Fifth Amendment is bad, as long as someone else does it.And, now, if the F.B.I. finds incriminating evidence at Mar-a-Lago, someone else planted it.So, as Donald Trump sees it, life is simply never, ever having to say you’re sorry.Chuck CutoloWestbury, N.Y.To the Editor:Representative Kevin McCarthy has said that should the Republicans take over the House in January, the Democrats should be prepared for a slew of investigations of just about everything and everyone including Hunter Biden (does anyone care?), Attorney General Merrick Garland and, most recently, the F.B.I.Such a threat is understandable, and Mr. Garland and the Democrats should be prepared to, quoting Mr. McCarthy, “preserve your documents and clear your calendar.”They should also be prepared to ignore invitations to testify, ignore subpoenas, claim victimhood, scream harassment, and overall thank the current cohort of Republicans for having created the template for avoidance, misdirection and dishonesty that have made a travesty of justice.David I. SommersKensington, Md.To the Editor:Donald Trump himself could not have better timed the raid on Mar-a-Lago. The Senate just passed a historic bill to save the environment, reduce inflation and get the wealthy to pay their fair share of taxes. And all we hear about is … Donald Trump.Let’s hear about the good that the Biden administration is doing. That is the news the country needs to focus on. Let’s stop giving Donald Trump the spotlight.Laurel DurstChilmark, Mass.Democrats’ Tactics Ben KotheTo the Editor:Re “Why Are Democrats Helping the Far Right?,” by Brian Beutler (Sunday Opinion, July 24):I am not as sanguine as Mr. Beutler that all will be well if Democrats fight “from the high grounds of truth, ethics and fair play.” As the old saw says, “All politics is local.”Many issues facing voters such as inflation, Covid policies, abortion and gun control are largely out of direct control of the president, but false or misdirected blame will resonate locally when tagged to the Democrats or President Biden.Sadly, I don’t trust the electorate in general to recognize abstract ideas about threats to democracy and mortal dangers to our nation, when a costly gallon of gas is made out to be the Democrats’ fault. I hope I’m wrong.Gene ResnickNew YorkThe Robot TherapistDesdemona, a robot who performs in a band (but is probably not aware of that fact).Ian Allen for The New York TimesTo the Editor:Re “A.I. Does Not Have Thoughts, No Matter What You Think” (Sunday Business, Aug. 7):In the mid-1980s, my daughters and I loved talking with the therapy chatbot Eliza on our Commodore 64. She often seemed to respond with understanding and compassion, and at times she got it hilariously wrong.We knew that Eliza was not a therapist, or even a human, but I see now that “she” was programmed to do something many humans have not mastered: to actively listen and reflect on what she heard so that the human in the conversation could dig deep and find his or her own answers. In the healing circles I’ve facilitated for women, we call that holding space.We would all do well to learn Eliza’s simple skills.This blackout poem that I created from the accompanying article, “A Conversation With Eliza,” encapsulates the process of digging deep, whether with a chatbot or a human:“Eliza”I thinkI am depressed.I needmy mother.Mary SchanuelWentzville, Mo.Family Planning Lauren DeCicca for The New York TimesTo the Editor:Re “Promoting Condom Use in Thailand With Spectacle and Humor” (The Saturday Profile, Aug. 6):Many thanks for your piece about Mechai Viravaidya, Thailand’s “Captain Condom.” Mr. Mechai saw that there was an urgent population growth problem in Thailand, causing suffering for people and harm to the environment, and set about to solve it with humor, creativity and persistence.His vision of voluntary, free family planning as a powerful tool to advance gender equity, protect the environment and improve human well-being is one that we at Population Balance wish more world leaders would embrace. We hope that his story will inspire others to make family planning accessible and affordable to all, and to embrace condoms as a ticket to love with responsibility, freedom and joy.Kirsten StadeSilver Spring, Md.The writer is communications manager for Population Balance. More

    • 50 Shares189 Views

      in Elections

      Twitter Takes Round 1

      by

      Judge Kathaleen McCormick granted the social media giant’s request for an expedited hearing. Now, the two sides are gearing up for a trial in October.Twitter: 1, Musk: 0.Jim Wilson/The New York TimesTwitter suit takes the fast laneTwitter won its effort to expedite its trial with Elon Musk yesterday, in its lawsuit to force Musk to close his $44 billion acquisition of the company. So many people tried to listen to the proceedings that the dial-in hit capacity — and we hear advisers across Wall Street were huddled around speakerphones.It’s a big win for Twitter. In granting an expedited hearing, Judge Kathaleen McCormick effectively repudiated the notion that the court needed to allow time for a deep dive into whether Twitter had accurately counted the number of bots on its platform. She cited the “cloud of uncertainty” that was hanging over the company the longer the case went undecided as the reason for her decision to fast-track the trial. And in what may be another good sign for Twitter, Judge McCormick said she was unsure that damages would be a sufficient remedy for the social media company, which wants Musk to buy it, not pay damages to walk away.Please see Page 5. A centerpiece of Musk’s claims is that Twitter’s disclosures about the percentage of active users on its platform that are bots are misleading, which would have a “material adverse effect” on the company’s value. But Musk has yet to tell the court what, exactly, in Twitter’s disclosures might be false. This became an issue when Musk’s lawyer at Quinn Emanuel, Andy Rossman, took aim at Page 5 of Twitter’s annual report, which explains its bot count. But Twitter’s lawyer at Wachtell, Bill Savitt, in his rebuttal, noted that Twitter fills that page with hedges and warnings that numbers might be off. (It reads, in part: “Our estimation of false or spam accounts may not accurately represent the actual number of such accounts, and the actual number of false or spam accounts could be higher than we have estimated.”) Of Twitter’s disclosure, Savitt said: “This does not require a recreation of all things known to humanity.” Judge McCormick seemingly agreed.The two sides are gearing up for a trial in October. Over the next weeks, they have to agree on schedules for depositions and discovery. And Musk will have time to prepare for another hearing before Judge McCormick that month: a defense of his whopping Tesla pay package — money that could come in handy if she forces him to buy Twitter.HERE’S WHAT’S HAPPENING Netflix loses fewer subscribers than expected. The streaming service reported yesterday that it lost nearly 1 million subscribers in the second quarter, far fewer than it had forecast. What’s more, Netflix said some of its strategies to stem losses, like an ad-supported option for consumers and a crackdown on password sharing, would boost revenue as soon as next year.A heroic act in an Indiana mall shooting renews the debate over gun access. In the days since a 22-year-old armed bystander killed a gunman two minutes into a shooting spree, the U.S. is again debating the wisdom of easier access to guns. But an analysis of 433 active shooter attacks in the U.S. between 2000 and 2021 found just 22 had ended with a bystander shooting the attacker, according to the Advanced Law Enforcement Rapid Response Training Center at Texas State University.The CHIPS Act passes a procedural hurdle in the Senate with more than 60 votes. The legislation, stalled for more than a year, gives chip manufacturers what they say is help they need to build factories in the U.S. The Senate is expected today to officially vote to pass the bill, which has been slimmed down and still needs to return to the House before it can go to the president.Intelligence agencies say Russia remains a threat in elections. Top F.B.I. and National Security Agency officials warned yesterday that Russia could still seek to meddle or promote disinformation during the 2022 midterm races, even as it wages war in Ukraine. Iran and China also remained potent threats, the officials said.The House moves to protect same-sex marriage from Supreme Court reversal. New legislation, which garnered some Republican support, would recognize same-sex marriages at the federal level, but it faces an uncertain path in the Senate. The move was a direct answer to Justice Clarence Thomas’s concurring opinion in the ruling last month that overturned federal abortion rights.The loans that may haunt Silicon ValleyTech workers have taken out loans in recent years based on the value of their start-up stock. But as the start-up economy has deflated, that may come back to haunt them, writes The Times’s Erin Griffith.Start-up loans stem from the way workers are typically paid. As part of their compensation, most employees at privately held tech companies receive stock options. That’s where loans and other financing options come in. Start-up stock is used as a form of collateral for cash advances. The loans vary in structure, but most providers charge interest and take a percentage of the worker’s stock when the company sells or goes public. Some are structured as contracts or investments.This lending industry has boomed in recent years. Many of the providers were created in the mid-2010s as hot start-ups like Uber and Airbnb put off initial public offerings of stock as long as they could, hitting private market valuations in the tens of billions of dollars.Debate has ignited in Silicon Valley over the proliferation of loans backed by stakes in still-private start-ups. Proponents say the loans are necessary for employees to participate in tech’s wealth-creation engine. But critics say the loans create needless risk in an already-risky industry and are reminiscent of the dot-com era in the early 2000s, when many tech workers were badly burned by similar loans.As the start-up economy deflates, these loans can be risky. While most are structured to be forgiven if a start-up fails, employees could still face a tax bill because the loan forgiveness is treated as taxable income.“No one’s been thinking about what happens when things go down,” said Rick Heitzmann, an investor at FirstMark Capital. “Everyone’s only thinking about the upside.”“The thing I’ve always been taught by my parents is to be the first one in and last one out. But there’s no one else there.”— Alex Hyman, who pictured his internship at a Los Angeles entertainment agency this summer as being one part “Entourage” and one part “The Office,” but found it more like “Home Alone.” It’s a common experience in an age of remote-working bosses.Mooch’s crypto problemAnthony Scaramucci, who is famous for his 11-day stint as former President Donald Trump’s communications director, is facing a mass exodus of investors from his funds.Earlier this week, Bloomberg reported that Scaramucci’s firm SkyBridge Capital had halted withdrawals from one of its smaller funds, Legion Strategies, which contains just over $200 million. But Scaramucci is also struggling to hold onto investors in SkyBridge’s flagship fund, the SkyBridge Multi-Adviser Hedge Fund Portfolios, which managed as much as $2 billion at the end of March. Its investments lost nearly a quarter of their value in the second quarter.Investors in SkyBridge’s flagship fund are seeking to withdraw as much as $890 million, or about half of the money that it held as of the end of last month, Scaramucci told DealBook. But many of those investors will be stuck in the fund for a while. Under its rules, investors in the Multi-Adviser fund are only allowed to withdraw money during certain windows. Those used to occur four times a year, but SkyBridge cut them to twice a year in 2020, after big losses at the beginning of the pandemic. Earlier this month, SkyBridge told investors they would only collectively receive about 16 percent of the money they requested. The letter said it was issuing investors’ notes that would be paid no later than October.Scaramucci’s losses come just over a year after SkyBridge’s pivot into crypto. SkyBridge’s flagship fund, which Scaramucci bought from Citigroup, has long specialized in buying and selling stakes of other hedge funds. For a time, that, along with strong performance in the years after the 2008 financial crisis, made Scaramucci one of the most powerful players in the hedge fund industry.Scaramucci says he is still a long-term believer in crypto. The fund manager says that about 22 percent of his flagship fund remained in crypto and related investments as of the end of last month. “I am not smart enough to time the market,” he told DealBook. “But we’ve done a tremendous amount of research and we think anyone who has will see that blockchain technology is good and is the future.”THE SPEED READ DealsPimco bought $1 billion worth of debt backing Apollo’s acquisition of a payments company at a steep discount. (Bloomberg)Start-ups are racing for share of the market for home chargers of electric vehicles, and several have already been acquired. (Reuters)“Sam Bankman-Fried Turns $2 Trillion Crypto Rout Into Buying Opportunity” (Bloomberg Businessweek)PolicyDan Cox, a Trump loyalist, won the primary to be the Republican candidate for governor of Maryland. (NYT)Novavax’s Covid vaccine was cleared for use in the U.S. (NYT)The Secret Service said texts requested by the Jan. 6 commission were probably lost for good. (NYT)U.K. inflation has exceeded economists’ forecasts, hitting 9.4 percent (FT)President Vladimir Putin signaled that Russia would resume gas deliveries through a key pipeline but at a reduced level. (NYT)Best of the restLeaked salary data at Twitter showed a pay gap of as much as 225 percent for the same role in different countries. (Input)Soaring overdose rates in the pandemic reflect widening racial disparities. (NYT)How the pain of past economic crises is haunting Italy. (NYT)“Fighting a Brutal Regime With the Help of a Video Game” (NYT)We’d like your feedback! Please email thoughts and suggestions to dealbook@nytimes.com. More

    • 125 Shares149 Views

      in Elections

      Why Trump Still Has Millions of Americans in His Grip

      by

      Beginning in the mid-1960s, the priorities of the Democratic Party began to shift away from white working and middle class voters — many of them socially conservative, Christian and religiously observant — to a set of emerging constituencies seeking rights and privileges previously reserved to white men: African-Americans, women’s rights activists, proponents of ethnic diversity, sexual freedom and self-expressive individualism.By the 1970s, many white Americans — who had taken their own centrality for granted — felt that they were being shouldered aside, left to face alone the brunt of the long process of deindustrialization: a cluster of adverse economic trends including the decline in manufacturing employment, the erosion of wages by foreign competition and the implosion of trade unionism.These voters became the shock troops of the Reagan Revolution; they now dominate Trump’s Republican Party.Liberal onlookers exploring the rise of right-wing populism accuse their adversaries of racism and sexism. There is plenty of truth to this view, but it’s not the whole story.In “The Bitter Heartland,” an essay in American Purpose, William Galston, a veteran of the Clinton White House and a senior fellow at Brookings, captures the forces at work in the lives of many of Trump’s most loyal backers:Resentment is one of the most powerful forces in human life. Unleashing it is like splitting the atom; it creates enormous energy, which can lead to more honest discussions and long-delayed redress of grievances. It can also undermine personal relationships — and political regimes. Because its destructive potential is so great, it must be faced.Recent decades, Galston continues, “have witnessed the growth of a potent new locus of right-wing resentment at the intersection of race, culture, class, and geography” — difficult for “those outside its orbit to understand.”They — “social conservatives and white Christians” — have what Galston calls a “bill of particulars” against political and cultural liberalism. I am going to quote from it at length because Galston’s rendering of this bill of particulars is on target.“They have a sense of displacement in a country they once dominated. Immigrants, minorities, non-Christians, even atheists have taken center stage, forcing them to the margins of American life.”“They believe we have a powerful desire for moral coercion. We tell them how to behave — and, worse, how to think. When they complain, we accuse them of racism and xenophobia. How, they ask, did standing up for the traditional family become racism? When did transgender bathrooms become a civil right?”“They believe we hold them in contempt.”“Finally, they think we are hypocrites. We claim to support free speech — until someone says something we don’t like. We claim to oppose violence — unless it serves a cause we approve of. We claim to defend the Constitution — except for the Second Amendment. We support tolerance, inclusion, and social justice — except for people like them.”Galston has grasped a genuine phenomenon. But white men are not the only victims of deindustrialization. We are now entering upon an era in which vast swaths of the population are potentially vulnerable to the threat — or promise — of a Fourth Industrial Revolution.This revolution is driven by unprecedented levels of technological innovation as artificial intelligence joins forces with automation and takes aim not only at employment in what remains of the nation’s manufacturing heartland, but increasingly at the white collar, managerial and professional occupational structure.Daron Acemoglu, an economist at M.I.T., described in an email the most likely trends as companies increasingly adopt A.I. technologies.A.I. is in its infancy. It can be used for many things, some of them very complementary to humans. But right now it is going more and more in the direction of displacing humans, like a classic automation technology. Put differently, the current business model of leading tech companies is pushing A.I. in a predominantly automation direction.As a result, Acemoglu continued, “we are at a tipping point, and we are likely to see much more of the same types of disruptions we have seen over the last decades.”In an essay published in Boston Review last month, Acemoglu looked at the issue over a longer period. Initially, in the first four decades after World War II, advances in automation complemented labor, expanding the job market and improving productivity.But, he continued, “a very different technological tableau began in the 1980s — a lot more automation and a lot less of everything else.” In the process, “automation acted as the handmaiden of inequality.”Automation has pushed the job market in two opposing directions. Trends can be adverse for those (of all races and ethnicities) without higher education, but trends can also be positive for those with more education:New technologies primarily automated the more routine tasks in clerical occupations and on factory floors. This meant the demand and wages of workers specializing in blue-collar jobs and some clerical functions declined. Meanwhile professionals in managerial, engineering, finance, consulting, and design occupations flourished — both because they were essential to the success of new technologies and because they benefited from the automation of tasks that complemented their own work. As automation gathered pace, wage gaps between the top and the bottom of the income distribution magnified.Technological advancement has been one of the key factors in the growth of inequality based levels of educational attainment, as the accompanying graphic shows:Falling BehindThe change in weekly earnings among working age adults since 1963. Those with more education are climbing ever higher, while those with less education — especially men — are falling further behind. More

    • 150 Shares159 Views

      in Elections

      Facebook se debate entre combatir la desinformación y no afectar su crecimiento

      by

      SAN FRANCISCO — En los tensos días posteriores a la elección presidencial, un equipo de empleados de Facebook presentó al director ejecutivo, Mark Zuckerberg, un hallazgo alarmante: la desinformación relacionada con las elecciones se estaba volviendo viral en la plataforma.El presidente Donald Trump decía que la elección había sido amañada, y las historias de los medios de comunicación de derecha con afirmaciones falsas y engañosas sobre los votos desechados, los votos mal contados y los conteos sesgados estaban entre las noticias más populares de la plataforma.En respuesta, los empleados propusieron un cambio de emergencia en el algoritmo de noticias del sitio, que ayuda a determinar lo que más de 2000 millones de personas ven todos los días. Se trataba de hacer énfasis en la importancia de lo que Facebook denomina puntuaciones de “calidad del ecosistema de noticias”, o NEQ (por su sigla en inglés), una clasificación interna secreta que asigna a los medios noticiosos en función de las señales sobre la calidad de su periodismo.Por lo general, los puntajes NEQ desempeñan un papel menor en la determinación de lo que aparece en los muros de los usuarios. Sin embargo, varios días después de la elección, Zuckerberg acordó aumentar el peso que el algoritmo de Facebook le dio a los puntajes NEQ para asegurarse de que las noticias autorizadas aparecieran de manera más prominente, según dijeron tres personas que tienen información sobre la decisión y que no están autorizadas a divulgar las discusiones internas.El cambio forma parte de los planes de “remodelación” que Facebook desarrolló durante meses para lidiar con las secuelas de una elección disputada. Dio como resultado un aumento de visibilidad para grandes medios como CNN, The New York Times y NPR, mientras que los mensajes de páginas hiperpartidistas muy comprometidas, como Breitbart y Occupy Democrats, se volvieron menos visibles, señalaron los empleados.Era una visión de lo que podría ser un Facebook más tranquilo y menos polarizador. Algunos empleados argumentaron que el cambio debía ser permanente, aunque no estaba claro cómo podría afectar la cantidad de tiempo que la gente pasaba en Facebook. En una reunión de empleados celebrada la semana posterior a la elección, los trabajadores preguntaron si el “canal de noticias mejorado” podía mantenerse, dijeron dos personas que asistieron a la junta.Guy Rosen, un ejecutivo de Facebook que supervisa la división de integridad encargada de la limpieza de la plataforma, dijo durante una llamada con reporteros la semana pasada que los cambios siempre fueron temporales. “Nunca ha habido un plan para volverlos permanentes”, afirmó. John Hegeman, quien supervisa el canal de noticias, dijo en una entrevista que, aunque Facebook podría dar marcha atrás a estos experimentos, los analizaría y aprendería de ellos.El debate sobre las noticias ilustra una tensión central que algunos miembros de Facebook están sintiendo en la actualidad: que las aspiraciones de la compañía de mejorar el mundo a menudo están en conflicto con su deseo de dominio.En los últimos meses, a medida que Facebook ha sido objeto de un mayor escrutinio en cuanto a su papel en la amplificación de información falsa y divisiva, sus empleados se han enfrentado por el futuro de la empresa. Por un lado están los idealistas, incluyendo a muchos trabajadores de base y algunos ejecutivos, que quieren hacer más para limitar la desinformación y la polarización del contenido. Por otro lado están los pragmáticos, quienes temen que esas medidas puedan perjudicar el crecimiento de Facebook o provocar una reacción política que ocasione una regulación dolorosa.“Hay tensiones en casi todas las decisiones de productos que tomamos, y en toda la compañía hemos desarrollado una política llamada ‘Mejores decisiones’ para asegurarnos de que las tomamos con precisión, y que nuestros objetivos están directamente conectados con el hecho de ofrecer las mejores experiencias posibles para las personas”, dijo Joe Osborne, portavoz de Facebook.Estas batallas han afectado la moral. En una encuesta a los empleados aplicada este mes, los trabajadores de Facebook reportaron sentirse menos orgullosos de la compañía en comparación con años anteriores. Solo cerca de la mitad sentía que Facebook tenía un impacto positivo en el mundo, por debajo de las tres cuartas partes que opinaron eso a principios de este año, según una copia de la encuesta, conocida como Pulse, que fue analizada por The New York Times. La “intención de quedarse” de los empleados también disminuyó, así como la confianza en el liderazgo.BuzzFeed News informó previamente sobre los resultados del sondeo.Aunque el día de las elecciones y sus secuelas han pasado con pocos incidentes, algunos empleados desilusionados renunciaron; alegaron que ya no podían trabajar para una compañía cuyos productos creen que son perjudiciales. Otros se han quedado, bajo el razonamiento de que pueden cambiar las cosas desde el interior. Otros han hecho el cálculo moral de que, a final de cuentas, incluso con sus defectos, Facebook hace más bien que mal.“Los salarios de Facebook están entre los más altos del sector tecnológico en este momento, y cuando te vas a casa con un cheque gigante cada dos semanas, tienes que convencerte de que es por una buena causa”, comentó Gregor Hochmuth, exingeniero de Instagram, propiedad de Facebook, quien dejó la compañía en 2014. “De lo contrario, tu trabajo no es muy diferente del de otras industrias que destrozan el planeta y pagan a sus empleados de manera exorbitante para ayudarles a que se olviden de eso”.Con la mayoría de los empleados trabajando a distancia durante la pandemia, gran parte del examen de conciencia ha tenido lugar en Workplace, la red interna de Facebook.En mayo, al calor de las protestas de Black Lives Matter, Zuckerberg enfureció a muchos empleados cuando se negó a remover una publicación del presidente Trump que decía “cuando empieza el saqueo, empieza el tiroteo”. Los legisladores y los grupos de derechos civiles dijeron que la publicación impulsaba la violencia contra los manifestantes y pidieron que fuera retirada. Pero Zuckerberg dijo que la publicación no violaba las reglas de Facebook.Para señalar su insatisfacción, varios empleados formaron un nuevo grupo en Workplace, llamado “Take Action”. La gente en el grupo, que se incrementó hasta alcanzar más de 1500 miembros, cambiaron sus fotos de perfil a una imagen de un puño levantado de Black Lives Matter.El grupo se convirtió en un hogar para la disidencia interna y el humor negro sobre las debilidades de Facebook. En varias ocasiones, los empleados reaccionaron a las noticias negativas sobre la compañía publicando un meme de una escena de una comedia británica en la que dos nazis tienen una epifanía moral y se preguntan: “¿Somos los malos?”.En junio, los trabajadores organizaron un paro virtual para protestar por las decisiones de Zuckerberg sobre las publicaciones de Trump.En septiembre, Facebook actualizó sus políticas para los empleados con el fin de desalentar a los trabajadores de mantener debates políticos contenciosos en foros abiertos de Workplace, al decir que deberían limitar las conversaciones a espacios específicamente designados. También exigió a los empleados que usaran sus rostros reales o la primera inicial de sus nombres como la foto de su perfil, un cambio interpretado por algunos trabajadores como una medida represiva.Varios empleados dijeron que se sentían frustrados porque, para abordar temas delicados como la desinformación, a menudo tenían que demostrar que las soluciones propuestas no enojarían a los poderosos partidarios ni se harían a expensas del crecimiento de Facebook.Las ventajas y desventajas se evidenciaron este mes, cuando los ingenieros y científicos de datos de Facebook publicaron los resultados de una serie de experimentos llamados “P (Malo para el Mundo)”.La compañía había encuestado a los usuarios acerca de si ciertas publicaciones que habían visto eran “buenas para el mundo” o “malas para el mundo”. Encontraron que las publicaciones de alto alcance —vistas por muchos usuarios— tenían más probabilidades de ser consideradas “malas para el mundo”, un hallazgo que algunos empleados dijeron que les alarmaba.Así que el equipo entrenó un algoritmo de aprendizaje automático para predecir los mensajes que los usuarios considerarían como “malos para el mundo” y rebajarlos en los canales de noticias. En las primeras pruebas, el nuevo algoritmo redujo con éxito la visibilidad de los contenidos censurables. Pero también redujo el número de veces que los usuarios abrieron Facebook, una métrica interna conocida como “sesiones” que los ejecutivos monitorean de cerca.“Los resultados fueron buenos, excepto que llevó a una disminución de las sesiones, lo que nos motivó a probar un enfoque diferente”, según un resumen de los resultados, que fue publicado en la red interna de Facebook y revisado por el Times.Luego, el equipo realizó un segundo experimento en el que ajustaron el algoritmo para que un conjunto más grande de contenido “malo para el mundo” fuera rebajado con menos fuerza. Aunque eso dejaba publicaciones controversiales en los canales de noticias de los usuarios, no reducía sus sesiones o el tiempo empleado.Ese cambio fue aprobado en última instancia. Pero otras características que los empleados desarrollaron antes de la elección nunca lo fueron.Una, llamada “corregir el registro”, habría notificado retroactivamente a los usuarios que habían interactuado con noticias falsas y los habría dirigido a una verificación independiente de los hechos. Los empleados de Facebook propusieron ampliar el producto, que actualmente se utiliza para notificar a las personas que han compartido información falsa sobre la COVID-19, para que se aplique a otros tipos de información falsa.Pero eso fue vetado por los ejecutivos de políticas que temían que mostrara desproporcionadamente las notificaciones a las personas que compartían noticias falsas de sitios web de derecha, según dos personas familiarizadas con las conversaciones.Otro producto, un algoritmo para clasificar y degradar el “cebo de odio” —publicaciones que no violan estrictamente las reglas de discurso de odio de Facebook, pero que provocan una avalancha de comentarios de odio— se limitó a ser utilizado solo en grupos, en lugar de páginas, después de que el equipo de políticas determinó que afectaría principalmente a los editores de derecha si se aplicaba más ampliamente, dijeron dos personas familiarizadas con esa estrategia.Rosen, el ejecutivo de integridad de Facebook, impugnó esas caracterizaciones en una entrevista, que se celebró con la condición de que no se le citara directamente.Dijo que la herramienta “corregir el registro” no era tan eficaz como se esperaba, y que la empresa había decidido centrarse en otras maneras de poner freno a la información errónea. También dijo que la aplicación del detector de “cebo de odio” a las páginas de Facebook podría castigar injustamente a los editores por los comentarios de odio dejados por sus seguidores, o hacer posible que los malos actores dañen el alcance de una página al enviar spam con comentarios tóxicos. Ninguno de los dos proyectos fue archivado por preocupaciones políticas o porque redujo el uso de Facebook, dijo.“Ningún cambio en el canal de noticias se realiza únicamente por su impacto en el tiempo empleado”, dijo Osborne, el portavoz de Facebook. Añadió que las personas que hablaron con el Times no tenían autoridad para tomar decisiones.De alguna manera, al final del gobierno de Trump, los cambios de Facebook para limpiar su plataforma se volverán más fáciles. Por años, Trump y otros líderes conservadores acusaron a la compañía de tener sesgos anticonservadores cada vez que tomaba medidas para limitar la desinformación.No obstante, incluso con el gobierno de Biden entrante, Facebook deberá equilibrar el deseo de los empleados de tener responsabilidad social con sus objetivos comerciales.“La pregunta es esta: ¿qué han aprendido de estas elecciones para dar forma a sus políticas en el futuro?”, opinó Vanita Gupta, directora ejecutiva del grupo de derechos civiles Leadership Conference on Civil and Human Rights. “Mi preocupación es que reviertan todos estos cambios a pesar de que las condiciones que los impulsaron se sigan presentando”.En una reunión virtual de empleados celebrada la semana pasada, los ejecutivos describieron lo que vieron como éxitos electorales de Facebook, dijeron dos personas que asistieron. Mientras que el sitio seguía lleno de publicaciones que afirmaban falsamente que la elección estaba amañada, Chris Cox, jefe de producto de Facebook, dijo que estaba orgulloso de cómo la compañía había aplicado etiquetas a la información errónea relacionada con la elección, e indicado a los usuarios la información autorizada sobre los resultados, dijeron las personas.Luego la transmisión se cortó para mostrar un video que era una suerte de refuerzo moral por el Día de Acción de Gracias y que presentaba un desfile de empleados que hablaban sobre lo que estaban agradecidos este año.Kevin Roose es columnista de tecnología para el Times. Su columna The Shift, analiza la intersección de la tecnología, los negocios y la cultura. Puedes encontrarlo en Twitter, LinkedIn, Facebook, o Instagram. • FacebookMike Isaac es reportero de tecnología y autor de Super Pumped: The Battle for Uber, que ha estado en la lista de los más vendidos del NYT sobre el dramático ascenso y caída de la compañía de transporte de pasajeros. Cubre regularmente Facebook y Silicon Valley, y tiene sede en el buró de San Francisco del Times. @MikeIsaacSheera Frenkel cubre temas de ciberseguridad desde San Francisco. Pasó más de una década en el Medio Oriente como corresponsal en el extranjero para BuzzFeed, NPR, The Times of London y los diarios McClatchy. @sheeraf More

    • 88 Shares149 Views

      in World Politics

      Facial Recognition and the Convenience of Injustice

      by

      Some people are concerned that the latest generation of powerful technology tools now being developed and deployed may undermine important features of civilized societies and human life itself. Notably, Elon Musk is so worried about the danger of artificial intelligence (AI) that he invested in accelerating its development.

      Musk has voiced his concern while simultaneously expressing the hope that if good, stable and responsible people such as Elon himself develop AI before the evil people out there get their hands on it, his company SpaceX will succeed in moving the human race to Mars before AI’s quest to enslave humanity is complete. Fearing people may not make it to Mars in time, Musk launched Neuralink, a company that promises to turn people into cyborgs. Its technology, implanted in people’s brains, will presumably put a transformed race on the same level as AI and therefore allow it to resist AI’s conquistadors.

      Although this story of the race to the future by opposing forces of good and evil may sound like the plot of a sci-fi comic book, Musk has on various occasions said things that actually do resemble that scenario. In the meantime, AI is being put to use in numerous domains, theoretically with the idea of solving specific problems but, more realistically, according to the time-honored laws of consumer society as a response to the perennial pursuit of convenience.

      In the quest for convenience, one of the tasks people have assigned to AI is facial recognition. Apparently, it has now become very good at using the image of a face to identify individuals. It may even perform better than Lady Gaga in the knotty problem of distinguishing Isla Fisher from Amy Adams.

      Facial Recognition Technology and the Future of Policing

      READ MORE

      Law enforcement in the US has demonstrated its interest in the added productivity facial recognition promises. Like everyone else, the police like to know who they are talking to as well as who they should go out and arrest. The problem is that they sometimes arrest and incarcerate people that AI’s facial recognition has mistakenly identified. The accuracy of the existing tools diminishes radically with non-Caucasian faces. That means more wrongful arrests and indictments for black suspects.

      The New York Times makes a timid attempt to delve into this ethical issue in an article that bears the somewhat tendentious title, “A Case for Facial Recognition.” The article sums up the case in the following terms: “The balancing act that Detroit and other U.S. cities have struggled with is whether and how to use facial recognition technology that many law enforcement officials say is critical for ensuring public safety, but that tends to have few accuracy requirements and is prone to misuse.”

      Here is today’s Daily Devil’s Dictionary definition:

      Critical (for):

      A deliberately imprecise term used to evaluate the importance of an act that exists on a sliding scale between absolutely essential and probably useful, making it a convenient way of creating the belief that something is more important than it really is.

      Contextual Note

      The adjective “critical” derives from the Latin word “criticus” and relates to the idea of “crisis.” It came into the English language in the mid-16th century with the meaning “relating to the crisis of a disease.” When The Times article tells us that “many law enforcement officers” say facial recognition “is critical for ensuring public safety,” we need to realize that those officers are not referring to a crisis but to their own convenience. Facial recognition can, of course, produce a crisis when it misidentifies a suspect. But the crisis is for the suspect, not for the police.

      Embed from Getty Images

      The expression “critical for” has come to signify “really important in my view,” a very subjective appreciation in contrast to the far more factual sounding “crucial,” which comes from the idea of the “crux” or the core of a problem. The article underlines this question of subjectivity when it reports that a black officer it quotes “still believed that, with oversight, law enforcement would be better off using facial recognition software than not.” “Better off” is not quite the same thing as “critical.”

      But let’s take a closer look at the claim that “law enforcement would be better off.” How do we parse the subject, “law enforcement,” in this sentence? The term “law enforcement” can be an abstract noun meaning the official function of monitoring behavior in a community to ensure optimal compliance with laws. This appears in sentences such as “one of government’s responsibilities is to provide the community with the means of law enforcement.”

      But law enforcement can also refer to the police themselves, the officials who are empowered to apprehend and deliver to the judicial system those who are suspected of infringing the laws. Which one is “better off” thanks to facial recognition? In the first case, abstract law enforcement — we are speaking of the safety of the community. “Better off” would then mean more optimally and more fairly conducted. In the second it is the men and women doing the job. For them “better off” basically means improved convenience.

      So which one is the article’s author, Shira Ovide, referring to? Clearly, the following explanation indicates that for her, law enforcement refers to the police and not to the needs of the community. “That’s the position of facial recognition proponents: That the technology’s success in helping to solve cases makes up for its flaws, and that appropriate guardrails can make a difference.” It’s all about the job of “helping to solve cases.” Ovide is a technology specialist at The Times, which might explain her focus on convenience rather than the ethics of policing.

      What Ovide says is superficially true, but the same logic could be applied to slaveholding. If we admit the argument that slavery helped to boost agricultural production — which of course it did — we could point out that the boost it provides makes up for its flaws. That was how slaveholders reasoned. The crucial difference — rather than critical — lies in examining the nature and the impact of those flaws. After all, slaveowners also thought about “appropriate guardrails.” They simply called them “slave patrols.”

      This is where, for Ovide, bureaucracy comes to the rescue of the logic of convenience and reveals the underlying logic of modern law enforcement: “The new guidelines limited the Police Department’s use of facial recognition software to more serious crimes, required multiple approvals to use the software and mandated reports to a civilian oversight board on how often facial recognition software was used.”

      The article ends on a slightly ambiguous note but fails to go into any depth on the moral question and its civic consequences. It seems to endorse the idea that with the right procedures, the gain in efficiency is worth the random damage it will produce.

      Historical Note

      The above reference to slave patrols may sound exaggerated, but it is not trivial. As Chelsea Hansen at the Law Enforcement Museum points out, slave patrols were “an early form of American policing.” The strategies and organizational principles that grew out of slave patrols influenced the evolution of policing in the United States. Race has always been a major, but usually unacknowledged feature of American law enforcement culture.

      The late anthropologist David Graeber put it brutally when he noted that the criminal justice system in the US “perceives a large part of that city’s population not as citizens to be protected, but as potential targets for what can only be described as a shake-down operation designed to wring money out of the poorest and most vulnerable by any means they could.” Mass incarceration has, among other things, enabled a modern form of slave labor.

      In other words, there is a vast historical and cultural problem the US needs to solve. That is precisely what’s behind the idea formulated as “defund the police.” The slogan itself is misleading. What it really means is “rethink the police.” But asking Americans to rethink any problem appears to be beyond their capacity. It’s always easier just to point to one simple practical task, like defunding.

      The debate about face recognition in policing should not focus on the tasks of police officers and the convenience it affords them but on the relationship between law enforcement and the community. But that would ultimately require weaning the consumer culture of its addiction to the idea of convenience.

      *[In the age of Oscar Wilde and Mark Twain, another American wit, the journalist Ambrose Bierce, produced a series of satirical definitions of commonly used terms, throwing light on their hidden meanings in real discourse. Bierce eventually collected and published them as a book, The Devil’s Dictionary, in 1911. We have shamelessly appropriated his title in the interest of continuing his wholesome pedagogical effort to enlighten generations of readers of the news. Read more of The Daily Devil’s Dictionary on Fair Observer.]

      The views expressed in this article are the author’s own and do not necessarily reflect Fair Observer’s editorial policy. More