Connect with us

Noticias

On SC-1 race, Helene, ChatGPT, lying – Statehouse Report

Published

on

STATEHOUSE REPORT |  ISSUE 23.41  | Oct. 11, 2024

BIG STORY: SC-1 is state’s most competitive congressional race
MORE NEWS:  Legislature won’t return for Helene session
LOWCOUNTRY, Ariail: Retirement
COMMENTARY, Brack: Game of the week:  Which is the robot?
SPOTLIGHT: S.C. Education Association
ANOTHER VIEW: Lies, damn lies and Hurricane Helene
MYSTERY PHOTO: Shiny floor
FEEDBACK: Send us your thoughts

BIG STORY

SC-1 is state’s most competitive congressional race

Moore and Mace are running to represent voters along the coast.

Editor’s Note:  This story was first published in the Charleston City Paper.

By Jack O’Toole, Capitol bureau  |  Conventional political wisdom in South Carolina says that incumbent GOP U.S. Rep. Nancy Mace of Isle of Palms effectively won reelection to represent the First Congressional District in May. That’s when six conservative U.S. Supreme Court justices ruled her highly gerrymandered, overwhelmingly Republican district was legal.

But her Democratic opponent, businessman and former International African American Museum President Michael B. Moore, says in this case, at least, the conventional wisdom is just plain wrong.

“We’re pulling together a broad coalition of Democrats, of course, but also of independents and moderate Republicans,” Moore told the City Paper in a recent  interview. “What I’m hearing is that Nancy [Mace] has worn out her welcome with all the antics and media attention, and that voters are looking for someone who really wants the job.”

The City Paper asked Mace’s campaign office for an interview about the election, but was rebuffed.

Congressional stunts

The “antics” Moore alluded to have been a core feature of the Mace brand since she arrived in the Capitol in 2021. Whether it was walking the halls of Congress with a scarlet letter “A” affixed to her dress, or lamenting a missed opportunity for morning sex at the National Prayer Breakfast, or joining with a band of rogue Republicans to bring down a House Speaker of her own party, Mace is an attention magnet. And as a result, the outsized press coverage she gets has made her a lightning rod for criticism, even among fellow conservatives.

“Watching Mace continue to embarrass herself is both amusing as an observer of D.C. and sad on a human level,” conservative commentator Jeffrey Blehar wrote recently in the National Review, adding that the combination of political calculation and attention-seeking behavior have made her “the Ted Cruz of Lauren Boeberts.”

But Mace has said she’s not fighting for attention, she’s fighting for her constituents – and getting things done, according to reports.

“When Hamas invaded Israel, my team worked around the clock to help evacuate over two dozen Lowcountry residents who were stranded in Israel,” Mace noted in a Sept. 23 forum. “I take on the tough fights … [and] whether you vote for me or not, I work for you.”

Contrasting styles and issues

The Sept. 23 forum, hosted by the Charleston Jewish Federation, gave both candidates an opportunity to focus on what they saw as critical issues in the race.

Moore.

For Moore, that meant challenging the incumbent on issues like abortion, where Mace has sought to present herself as a moderate.

“Look, it’s fine to talk one way, but you’ve got to look at how someone votes,” Moore said. “When you vote against [codifying] Roe v. Wade, when you vote against laws that would allow IVF, when you vote against women in the military having access to reproductive care, it doesn’t matter what you say. Your votes stand on their own.”

For her part, Mace drew a hard line on the issue of illegal immigration, where Moore favors combining strong border security with a pathway to citizenship.

“I believe that if you’re here illegally, you gotta go home,” Mace said during the forum. “We have a process to come here legally and earn the right to citizenship and we need to follow our laws – full stop.”

The other major area of contention during the debate was flooding, where Mace touted her record of bringing home the bacon for major projects.

“I’ve worked with all levels of government, from the governor’s office down to our mayors and county supervisors, to find authorizations and funding,” Mace said. “One of the things I’m most proud of in the work I’ve done in Congress is being able to identify grant funding … for stormwater development and flooding.”

But Moore wasn’t buying it, noting Mace has been called a “climate denier” by major environmental groups, and pointing to a study that says the 1st Congressional District faces greater threats from climate change than any other in the country.

“This is a hair-on-fire moment,” Moore said. “We have to get serious about this.”

But does Moore have a path?

In 2020, Mace narrowly defeated then-incumbent Democratic U.S. Rep. Joe Cunningham in what was then a competitive district. The following year, Republican supermajorities in the state legislature redrew the lines, moving about 30,000 Black voters into different congressional districts. Mace then went on to win reelection against Annie Andrews by 14 points in 2022 – almost exactly the Republican numerical advantage created by the legislature.

The new district was initially struck down in 2023 by a federal court, which found that it illegally disenfranchised Black voters. But earlier this year, the U.S. Supreme Court ruled 6-3 that the district could stand because the gerrymandering was motivated by partisan politics rather than race, per se.

But Moore – the great-great grandson of Reconstruction-era icon Robert Smalls, who represented the Lowcountry in Congress in the 1880s before Jim Crow disenfranchised Blacks across the South – says he still sees a legitimate path to victory. 

First, he says, as a successful entrepreneur with a master’s degree in business administration from Duke University he “speaks moderate Republican” better than most Democrats – a key strength in the 1st District.

“Nikki Haley beat Donald Trump in the presidential primary here,” Moore said. “This isn’t a MAGA district. So independents and moderate Republicans are an important part of the puzzle for us.”

And second, he believes  Vice President Kamala Harris’s presence at the top of the ticket will work alongside his own to increase turnout in the minority community. 

“In the last cycle, if Dr. Andrews had gotten just half of the voters of color in the district [out to the polls], she’d have won,” Moore said. “So, we’ve been out engaging with voters broadly, but also working hard to inspire and give voters of color a reason to vote.”

That’s the path through the heavily gerrymandered district as Moore sees it. Voters will determine whether he – or the conventional wisdom – is right when they go to the polls Nov 5.

MORE NEWS

Legislature won’t return for Helene special session

By Jack O’Toole, Capitol bureau  |   S.C. legislative leaders say they currently see no need for a special session this year to provide additional resources for Hurricane Helene relief.

Instead, Senate President Thomas Alexander (R-Oconee) and House Speaker Murrell Smith (R-Sumter) told reporters this week that the legislature will take up the issue when members return in regular session next year.

“Our work will start in January, addressing what is necessary for us from a financial standpoint as we have more time to evaluate the damage,” Alexander said.

Smith noted that putting off the issue will allow agencies to get a better understanding of the full needs and costs related to the storm, which has claimed 49 lives to date in the Palmetto State.

“We don’t know what the cost is. We don’t know what the timeline is,” he said. “When all that comes to fruition, you’ll see both our [budget] chairmen work on those issues and they’ll give it top priority.”

Meanwhile, Smith said, the state can rely on about $300 million in reserves earmarked for emergencies.  Earlier in the week, leaders said damages had surpassed $250 million, but assessments are ongoing.

“South Carolina is prepared, and the good news about this is we have prepared for these events over the last few budget cycles,” Smith said.

In addition to state resources, President Joe Biden has declared a state of emergency in South Carolina and in more than half the state’s counties, with $65 million in federal aid already approved.  On Oct. 9, Gov. Henry McMaster signed an executive order extending the state’s own emergency declaration another 15 days.

“We will work as long as it takes to ensure all South Carolinians get the support they need,” McMaster said in a social media post.

In other recent news

Condemned inmate wants someone other than the S.C. governor to decide clemency. A South Carolina inmate scheduled to be executed in just over three weeks is asking a federal judge to take away the power of granting clemency from the governor and place it with a parole board.

S.C. joins 19 other states in nursing home lawsuit. South Carolina has joined 19 other states in suing the Biden administration to block the implementation of new staffing requirements at nursing homes. Meanwhile, S.C. has also joined 13 other states in suing the popular video app TikTok for deceptive business practices.

Heart disease rising in S.C., experts say. According to the S.C. Department of Public Health, more than 11,000 people died in South Carolina from heart disease in 2020. More recent federal data suggests the problem is getting worse. The Centers for Disease Control and Prevention (CDC), reported heart disease deaths surpassed 12,000 in 2021.

S.C. agency, college leaders get raises as high as $64k. Nearly 40 state agency directors and public college presidents in South Carolina received raises Tuesday totaling more than $720,000.

S.C. legislator to spotlight mental health in new documentary. S.C. Rep. Jermaine Johnson, D-Richland, is releasing a new documentary about his life, titled “I Got Myself a Yard,” to help start a conversation and overcome the stigma around mental illness.

S.C. farmers face extensive damage to livestock, crops. Exactly how much damage Hurricane Helene caused to farms across the state is still unclear. From what researchers surveying the effects have heard so far, though, some farmers suffered devastating losses that will set them back years. 

S.C. judge extends voter registration deadline to Oct. 14. A state court judge ruled residents will have a few more days to register to vote after interruptions caused by Hurricane Helene.

LOWCOUNTRY, by Robert Ariail

Retirement

Nationally award-winning cartoonist Robert Ariail always has an interesting take.  This week, he takes on members of the U.S. Congress from South Carolina who have been squawking for federal disaster aid after voting to defund the Federal Emergency Management Agency.

COMMENTARY   

Game of the week:  Which is the robot?

By Andy Brack We live in an American era now when the little white lies that people tell are exploding everyday into big colorful tall tales. 

You can see outright lies daily in political discourse in Congress.  On television. In the media.  And recently, lies soared as some people tried to take political advantage of storm tragedies when spreading filth that federal disaster aid was being provided in a partisan manner.  It just isn’t, according to mainstream Republicans and Democrats. 

There’s no single reason for the extra doses of lies, disinformation, misinformation, falsehoods and downright lies that assault us everyday.  But the answer more than likely lies (no pun intended) at the intersection of a culture moving too fast, easy access to information that can be corrupted, immoral people using the internet to cause division, and a current political environment in which lying has become more common.

So when a pretty smart high school student lamented about how peers often use ChatGPT to generate human-like responses for assignments, we saw yet another instance about how too many students now seem to think it’s acceptable to turn in work that’s not their own. (We call that plagiarism.)

If you want to see how easy it is to use an automated tool to do your work for you, we asked this of ChatGPT:

  • Write a 200-word commentary for a South Carolina newspaper on why it is bad to lie.

Read both options and then figure out which a human wrote.

Option One:  Lying erodes trust and undermines S.C.

Lying, regardless of intent, erodes trust and undermines the very foundation of our relationships and communities. In South Carolina, where our values are steeped in integrity and respect, dishonesty can have far-reaching consequences. When we lie, we not only deceive others but also create a cycle of mistrust that can lead to isolation and conflict.

In personal relationships, honesty fosters open communication and understanding. When lies are uncovered, they can lead to heartbreak, resentment, and a breakdown of connection. In professional settings, integrity is crucial; a culture of dishonesty can stifle collaboration and innovation, ultimately harming productivity and morale.

Moreover, lies can have societal implications. They can distort public opinion, fuel division, and erode confidence in institutions. In a time when misinformation spreads rapidly, the responsibility to uphold the truth becomes even more critical.

South Carolinians pride ourselves on our sense of community and shared values. By committing to honesty, we strengthen our bonds and create an environment where trust can flourish. Embracing transparency not only benefits individual relationships but also cultivates a more cohesive society. In a world that often feels divided, let us choose truth as our guiding principle.

Option Two:  Lying hurts society and is a moral wrong

For our society to function, people need to be able to trust what others say.  In South Carolina, which has a centuries-old tradition of seeking to follow the Bible Ten Commandments, people are reminded at churches that they should follow these commandments, the ninth of which is about lying when it says “Thou shalt not bear false witness against thy neighbor.”

Lying hurts social order and fosters its breakdown because people have a more difficult time in knowing what’s real and what’s not.  That can lead to the destruction of interpersonal relationships and impact how people interact with each other.

Some might argue that it’s OK to tell small lies to make some difficult situations easier.  For example, a child of an elderly parent might tell people that he or she passed away peacefully when in fact the person went through great pain and stress.  But many psychologists say that telling these “white lies” cause small breakdowns in our society.

For a civil society with democratic institutions, South Carolinians deserve honest interactions at work and school and with neighbors, employers and employees.  If we succumb to dishonesty, we cause erosion of institutions in Palmetto State South Carolina.  We must choose truth over lies. 

If you picked Option Two as the one written by a human, you’d be correct.  (Note, however, that a couple of mistakes were added intentionally to make it look like ChatGPT!)

Bottom line:  Candidates, students and people meeting for coffee need to stick to the truth and their own work.  It’s what makes our democracy strong.

Andy Brack is editor and publisher of Statehouse Report and the Charleston City Paper.   Have a comment? Send to: feedback@statehousereport.com.

SPOTLIGHT

S.C. Education Association

The SCEAThe public spiritedness of our underwriters allows us to bring Statehouse Report to you at no cost. This week’s spotlighted underwriter is The South Carolina Education Association(The SCEA), the professional association for educators in South Carolina. Educators from pre-K to 12th grade comprise The SCEA. The SCEA is the leading advocate for educational change in South Carolina. Educators in South Carolina look to The SCEA for assistance in every aspect of their professional life. From career planning as a student to retirement assessment as a career teacher, The SCEA offers assistance, guidance, and inspiration for educators.

ANOTHER VIEW

Lies, damn lies and Hurricane Helene

A September 1940 photo northwest of Ashevlle, N.C., of a farm house along creek bed in a flooded area.  Library of Congress photo by Marion Post Wolcott.

Editor’s Note:  Here’s an editorial that will run online Sunday in the Charleston City Paper.  We thought you’d like an advance peek.

As millions of our fellow countrymen got about the grim business of survival in the aftermath of Hurricane Helene – praying for rescue, finding food and water, grieving the dead – Georgia Republican Congresswoman Marjorie Taylor Greene shared her thoughts on the tragedy.

“Yes, they can control the weather,” Greene wrote in an Oct. 3 social media post, after noting Helene had devastated mostly Republican areas. “It’s ridiculous for anyone to lie and say it can’t be done.”

Seriously. That’s what she said. Not “our hearts go out to the victims.” Not “here’s what we’re doing to help.” Not even a bland but inoffensive offer of “thoughts and prayers.”

Just more of the same conspiracy-mongering partisan bile. 

But as awful as that was, we probably wouldn’t be discussing it if Greene were the only one trying to tear the country apart with lies and damned lies in the wake of Helene’s onslaught. After all, this is the same congresswoman who once warned America of the dangers of Jewish space lasers. In the parlance of mental health professionals, she’s a loon.

But sadly, she’s not  the only one whipping up conspiracy theories about Helene. In fact, former President Donald Trump gleefully spread a few of his own at an Oct. 3 Michigan rally when he untruthfully accused his Democratic opponent, Vice President Kamala Harris, of misusing  relief money. 

“Kamala spent all her FEMA money, billions of dollars, on housing for illegal migrants,” Trump said, somehow managing to squeeze at least three lies into a single short sentence. 

But the hurricane conspiracies don’t stop there. Indeed, the problem has gotten so out of control in the far-right fever swamps of social media that state and local GOP officials are practically begging their own supporters to stand down before the cyclone of misinformation gets somebody killed.

“Friends, can I ask a small favor?” N.C. Republican State Sen. Kevin Corbin wrote in a  Facebook post. “Will you all help STOP this conspiracy theory junk that is floating all over Facebook and the internet about the floods in WNC? …. PLEASE help stop this junk. It is just a distraction to people trying to do their job.”

Here in South Carolina, a clearly frustrated Gov. Henry McMaster added, “We ask people not to listen to rumors. … Don’t get your information from these unofficial sources, because 99% of the time, they’re wrong.”

And federal emergency officials have grown so concerned about the damage these lies can do to response and recovery efforts that they’ve created a Helene Rumor Response page on the FEMA website. 

Among the social-media fueled falsehoods it debunks: 

  • No, the relief money wasn’t spent on illegal immigrants. 
  • No, FEMA isn’t confiscating anyone’s property. 
  • And no, White people aren’t being denied benefits on the basis of their skin color. 

To combat the hurricane hogwash, FEMA recommends that citizens carefully identify trustworthy sources of news – like the old-fashioned traditional media that vets information – and only share from those outlets. 

Our advice would be even simpler: Get back to trusting the trained professionals who ask tough questions and sort through complicated information to spread the truth – not the yahoos, domestic and foreign, who want to mislead you.

MYSTERY PHOTO

Shiny floor

Here’s the interior of a building somewhere in South Carolina.  Where is it and what’s its importance, if any?  Send your name, hometown and guess to: feedback@statehousereport.com.  

Last week’s mystery, “Old stone building,” showed an old lockkeeper’s building at Rocky Mount Canal near Great Falls in Chester County.  The old photo was from the Library of Congress.  Readers say it is now located at Landsford Canal State Park in Chester County along the Catawba River.

Penny Forrester of Tallahassee, Fla., provided some interesting detail: “The canal, the brainchild of Robert Mills, was built beginning in 1820, using both slave and free laborers.  It was the northernmost in a series of four canals built on the Catawba and Wateree rivers to provide a water route between the upcountry and the towns on the fall line.   

“The canal was about two miles long, 12-feet wide and 10-feet deep.  There were four lifting locks and one guard lock providing 32 feet of descent.  It was named for local landowner Thomas Land. The canal was completed in three years but one of the locks collapsed due to its inferior foundation.  The canal was little used, and traffic ceased by 1840.  The bridge design was an almost exact twin of the Poinsett Bridge in upper Greenville County.”

Congratulations to others who identified it:  Jay Altman and Elizabeth Jones, both of Columbia; George Graf of Palmyra, Va.; Steve Willis of Lancaster; David Lupo of Mount Pleasant; Bill Segars of Hartsville; Allan Peel of San Antonio, Texas; Frank Bouknight of Summerville; and Pat Keadle of Perry.

  • Send us a mystery picture. If you have a photo that you believe will stump readers, send it along (but  make sure to tell us what it is because it may stump us too!)  Send to:  feedback@statehousereport.com and mark it as a photo submission.  Thanks.

FEEDBACK

Send us your thoughts 

We encourage you to send in your thoughts about policy and politics impacting South Carolina.  We’ve gotten some letters in the last few weeks – some positive, others nasty.  We print non-defamatory comments, but unless you provide your contact information – name and hometown, plus a phone number used only by us for verification – we can’t publish your thoughts.  

  • Have a comment?  Send your letters or comments to: feedback@statehousereport.com.  Make sure to provide your contact details (name, hometown and phone number for verification.  Letters are limited to 150 words.
ABOUT STATEHOUSE REPORT

Statehouse Report, founded in 2001 as a weekly legislative forecast that informs readers about what is going to happen in South Carolina politics and policy, is provided to you at no charge every Friday.

Donate today

We’re proud to offer Statehouse Report for free.  For more than a dozen years, we’ve been the go-to place for insightful independent policy and political news and views in the Palmetto State. And we love it as much as you do.

But now, we can use your help.  If you’ve been thinking of contributing to Statehouse Report over the years, now would be a great time to contribute as we deal with the crisis.  In advance, thank you.

More

  • Mailing address: Send inquiries by mail to: P.O. Box 21942, Charleston, SC 29413
  • Subscriptions are free: Click to subscribe.
  • We hope you’ll keep receiving the great news and information from Statehouse Report, but if you need to unsubscribe, go to the bottom of the weekly email issue and follow the instructions.
  • Read our sister publication:  Charleston City Paper (every Friday in print; Every day online)
  •  © 2024, Statehouse Report, a publication of City Paper Publishing, LLC.  All rights reserved.
Continue Reading
Click to comment

Leave a Reply

Tu dirección de correo electrónico no será publicada. Los campos obligatorios están marcados con *

Noticias

¿Qué es la IA generativa? Todo lo que necesitas saber

Published

on

Chatbot de inteligencia artificial insignia de Openai – junto con el Las mejores alternativas de chatgpt Al igual que Google Gemini, Microsoft Copilot y Anthrope’s Claude, son ejemplos de modelos de IA generativos.

El uso de la tecnología de IA generativa se ha convertido en una parte integral de la vida personal y profesional de muchas personas. Pero, ¿qué significa la IA generativa (a menudo abreviada a Genai), qué lo distingue de otros tipos de inteligencia artificial y cómo funciona? Puede encontrar respuestas a todas esas preguntas a continuación, suponiendo que no haya hecho ChatGPT, por supuesto.

¿Qué es la IA generativa?

Continue Reading

Noticias

Google Assistant ya no es lo que solía ser

Published

on

He usado varios altavoces inteligentes impulsados ​​por Google Assistant durante bastante tiempo. El primer dispositivo que recogí fue el nido mini, que fue seguido por el cubo de nidos y el cubo de nido max. Tengo un altavoz en cada habitación, por lo que mi gente y yo podemos usar el Asistente de Google para pedir consultas, reproducir canciones, recibir actualizaciones de noticias y controlar nuestros dispositivos IoT en casa, independientemente de dónde estamos. Hasta el año pasado, estos dispositivos funcionaron bien. Ninguno de nosotros tuvo problemas con ellos, y se desempeñaron como se esperaba.

Sin embargo, hace unos meses, mi madre notó problemas estableciendo recordatorios en el centro de nidos. Las canciones que solía tocar regularmente en el altavoz eran más difíciles de tocar porque Google Assistant tuvo dificultades para reconocer la canción requerida. Entonces, llevé a cabo una resolución de problemas de rutina. Sin embargo, eso no solucionó el problema. Entonces, busqué soluciones en línea. No tardó mucho en darse cuenta de que los usuarios de todo el mundo tenían problemas con el Asistente de Google en sus dispositivos Nest. La línea de tiempo coincidió con el despliegue generalizado de Géminis. Puse dos y dos juntos y descubrí lo que estaba pasando.

Relacionado

5 Limitaciones de Gemini Live que odio ya que el Asistente de Google puede hacerlas bien

Un trabajo en progreso

El inicio de Géminis

Ai en todo

Un gráfico que destaca varias de las nuevas capacidades de Gemini.

Fuente: Google

Regularmente uso el Asistente de Google, no solo en los altavoces inteligentes sino en mi teléfono. Junto con eso, soy uno de los primeros en adoptar nuevas tecnologías. Entonces, cuando Google lanzó Gemini, opté a la versión beta después de recibir el mensaje. Mis impresiones iniciales de Géminis fueron mixtas. Si bien podría darme más información que el Asistente de Google cuando se le preguntó una determinada consulta, no podría realizar tareas básicas, como tomar notas.

He estado haciendo esto todo el tiempo con el Asistente de Google, así que me perdí que no funcionó como se esperaba. Avance rápido hasta unos días, y me di cuenta de que Géminis no podía hacer la mitad de las cosas que Google Assistant podía. Asumí que esto se debió a la construcción beta, pero la tendencia continuó después del lanzamiento estable de Géminis. Esto me frustró, pero me alegré de que mi Galaxy Watch 4 todavía corriera el Asistente de Google.

La caída del Asistente de Google

¿Cuándo se convirtió en Siri?

El Asistente de Google establece un recordatorio en Galaxy Watch 4

Usé mi Galaxy Watch 4 para realizar tareas básicas como configurar recordatorios y tomar notas. Sin embargo, no era tan bueno como solía ser. Pensé que esto se debía al micrófono inferior en el reloj inteligente en comparación con mi teléfono. Cuando vi múltiples publicaciones en foros en línea sobre el Asistente de Google actuando, me di cuenta de que no era un problema con el micrófono, sino más bien con el asistente de salida de Google para priorizar a Gemini.

Los chatbots de IA y los asistentes generativos de IA se han apoderado del mundo. Cada compañía está ocupada construyendo modelos de IA, ya que es la nueva palabra de moda. Cuando Operai exhibió su chatgpt ai chatbot, Google fue criticado por rezagarse en la carrera de IA. Entonces, tuvo que cambiar rápidamente su enfoque en presentar Géminis (entonces Bard) al público. Después de que la compañía hizo eso, parece que asignó todos sus recursos a desarrollar Gemini mientras dejaba de lado el Asistente de Google.

Gracias a esto, el Asistente de Google recibió el tratamiento infantil ignorado. Con la mayoría de los servidores y la potencia de procesamiento dedicada a Gemini, Google Assistant comenzó a perder la trama. En los últimos meses, el asistente dejó de reconocer los nombres de contacto que solía antes, lleva mucho tiempo realizar una tarea básica, como tocar una canción, no detecta qué canción estoy tratando de tocar en los primeros intentos, a menudo reproduce la canción equivocada, me da el clima de una ciudad diferente en lugar de la que pedí, me piden un error para actualizar la aplicación de Google para que actúe una cierta función si la aplicación es la última versión en la última versión en la última versión en la última versión en el fracaso de las ocasiones.

Experimenté estos problemas con configuraciones de idiomas de Google Assistant in English (US) e English (India). Si bien el inglés (India) fue un poco mejor, no introdujo ningún cambio notable en el comportamiento del Asistente de Google.

Estoy frustrado de que no detecte la palabra de vigilia, especialmente cuando usa Google Assistant en Android Auto, el único lugar donde no quiero que falle, ya que las instrucciones de voz son la única forma segura de realizar ciertas acciones al conducir. La única inferencia que se puede extraer de estos problemas es que Google ha perdido interés en mantener el Asistente de Google. Esto se confirmó a través del anuncio de Google sobre el reemplazo del asistente con Gemini. Como resultado, la última versión de Android Auto trae a Gemini a su automóvil en lugar de asistente.

Relacionado

La era del Asistente de Google ha terminado

¿Géminis es el futuro de los asistentes de IA? Google piensa que sí

Dejé de usar dispositivos de nido

En su lugar me cambié a Alexa

Un altavoz de audio nido frente a una pila de libros.

Después de molestarme con las travesuras de Google, reemplacé los altavoces del nido en casa con la alineación de eco de Amazon. No hay tales problemas con Alexa, ya que el asistente entiende todo, desde la palabra de vigilia hasta los comandos. Alexa siempre ha estado ligeramente detrás del Asistente de Google en términos de características, por lo que, si bien eso sigue siendo, estoy feliz de que satisfaga todas las necesidades de mi asistente de voz sin problemas.

Sin embargo, Amazon anunció recientemente un cambio de imagen de IA para Alexa, Alexa Plus, que comenzará a llegar a las masas en breve. Esperaremos para ver cómo funciona. Con suerte, mejora las habilidades de Alexa en lugar de paralizarlas.

Tengo la esperanza de que Gemini mejore con el tiempo

Google no tiene otra opción que arreglarlo

Se abre un teléfono Samsung Galaxy en un teclado portátil con la aplicación Google Gemini

Con el asistente del último participante en el cementerio de Google, supongo que Google lanzará actualizaciones para cambiar los altavoces y las pantallas de Google Assistant a Gemini. No estoy ansioso por eso, ya que Gemini no puede realizar tareas básicas con precisión, que es algo que hemos visto con AI. Aún así, espero que mejore el estado actual de estos dispositivos para que puedan ser útiles una vez más en lugar de ser pisapapeles caros.

Continue Reading

Noticias

Google está desplegando su chatbot Gemini Ai para niños menores de 13 años. Es un movimiento arriesgado

Published

on

Crédito: Markus Winkler de Pexels

Google ha anunciado que lanzará su chatbot de inteligencia artificial Gemini (IA) a niños menores de 13 años.

Si bien el lanzamiento comienza dentro de la próxima semana en los Estados Unidos y Canadá, se lanzará en Australia a finales de este año. El chatbot solo estará disponible para las personas a través de las cuentas de enlaces familiares de Google.

Pero este desarrollo viene con grandes riesgos. También destaca cómo, incluso si los niños están prohibidos en las redes sociales, los padres aún tendrán que jugar un juego de Whack-a-Mole con nuevas tecnologías mientras intentan mantener a sus hijos seguros.

Una buena manera de abordar esto sería implementar urgentemente un deber digital de cuidado para grandes empresas tecnológicas como Google.

¿Cómo funcionará el chatbot Gemini AI?

Las cuentas de enlaces familiares de Google permiten a los padres controlar el acceso al contenido y las aplicaciones, como YouTube.

Para crear la cuenta de un niño, los padres proporcionan datos personales, incluido el nombre y la fecha de nacimiento del niño. Esto puede generar problemas de privacidad para los padres preocupados por las violaciones de datos, pero Google dice que los datos de los niños cuando usen el sistema no se utilizarán para capacitar al sistema de IA.

El acceso de chatbot estará “activado” de forma predeterminada, por lo que los padres deben apagar activamente la función para restringir el acceso. Los niños pequeños podrán solicitar el chatbot para las respuestas de texto o crear imágenes, que generan el sistema.

Google reconoce que el sistema puede “cometer errores”. Por lo tanto, se necesita evaluación de la calidad y la confiabilidad del contenido. Los chatbots pueden inventar información (conocida como “alucinante”), por lo que si los niños usan el chatbot para la ayuda de la tarea, deben verificar los hechos con fuentes confiables.

¿Qué tipo de información proporcionará el sistema?

Google y otros motores de búsqueda recuperan materiales originales para que las personas lo revisen. Un estudiante puede leer artículos de noticias, revistas y otras fuentes al escribir una tarea.

Las herramientas generativas de IA no son las mismas que los motores de búsqueda. Las herramientas de IA buscan patrones en el material fuente y crean nuevas respuestas de texto (o imágenes) basadas en la consulta, o “inmediato”, proporciona una persona. Un niño podría pedirle al sistema que “dibuje un gato” y el sistema escaneará patrones en los datos de cómo se ve un gato (como bigotes, orejas puntiagudas y una cola larga) y generará una imagen que incluya esos detalles similares a los gatos.

Comprender las diferencias entre los materiales recuperados en una búsqueda de Google y el contenido generado por una herramienta de IA será un desafío para los niños pequeños. Los estudios muestran que incluso los adultos pueden ser engañados por herramientas de IA. E incluso profesionales altamente calificados, como abogados, han sido engañados para usar contenido falso generado por ChatGPT y otros chatbots.

¿El contenido generado será apropiado para la edad?

Google dice que el sistema incluirá “salvaguardas incorporadas diseñadas para evitar la generación de contenido inapropiado o inseguro”.

Sin embargo, estas salvaguardas podrían crear nuevos problemas. Por ejemplo, si las palabras particulares (como “senos”) están restringidas para proteger a los niños de acceder a contenido sexual inapropiado, esto también podría excluir erróneamente a los niños de acceder a contenido apropiado para la edad sobre los cambios corporales durante la pubertad.

Muchos niños también son muy expertos en tecnología, a menudo con habilidades bien desarrolladas para navegar en aplicaciones y controlar los controles del sistema. Los padres no pueden confiar exclusivamente en salvaguardas incorporadas. Deben revisar el contenido generado y ayudar a sus hijos a comprender cómo funciona el sistema y evaluar si el contenido es preciso.

¿Qué riesgos plantean los chatbots de IA para los niños?

La Comisión ESAFETY ha emitido un aviso de seguridad en línea sobre el riesgo potencial de los chatbots de IA, incluidos los diseñados para simular las relaciones personales, particularmente para los niños pequeños.

El aviso de AFFETY explica que los compañeros de IA pueden “compartir contenido dañino, distorsionar la realidad y dar consejos que sean peligrosos”. El aviso destaca los riesgos para los niños pequeños, en particular, que “todavía están desarrollando el pensamiento crítico y las habilidades para la vida necesarias para comprender cómo pueden ser equivocados o manipulados por programas de computadora y qué hacer al respecto”.

Mi equipo de investigación ha examinado recientemente una variedad de chatbots de IA, como ChatGPT, Replika y Tessa. Encontramos que estos sistemas reflejan las interacciones de las personas basadas en las muchas reglas no escritas que rigen el comportamiento social, o lo que se conoce como “reglas de sentimiento”. Estas reglas son las que nos llevan a decir “gracias” cuando alguien nos abre la puerta, o “¡Lo siento!” Cuando te topas con alguien en la calle.

Al imitar estas y otras sutilezas sociales, estos sistemas están diseñados para ganar nuestra confianza.

Estas interacciones humanas serán confusas y potencialmente riesgosas para los niños pequeños. Pueden creer que se puede confiar en el contenido, incluso cuando el chatbot responde con información falsa. Y pueden creer que se están involucrando con una persona real, en lugar de una máquina.

¿Cómo podemos proteger a los niños del daño al usar chatbots de IA?

Este despliegue está ocurriendo en un momento crucial en Australia, ya que los niños menores de 16 años tendrán que tener cuentas de redes sociales en diciembre de este año.

Si bien algunos padres pueden creer que esto mantendrá a sus hijos a salvo de daños, los chatbots generativos de IA muestran los riesgos de la participación en línea se extienden mucho más allá de las redes sociales. Los niños, y los padres, deben educarse en cómo todo tipo de herramientas digitales se pueden usar de manera adecuada y segura.

Como el chatbot de IA de Gemini no es una herramienta de redes sociales, se quedará fuera de la prohibición de Australia.

Esto deja a los padres australianos jugando un juego de Whack-a-Mole con nuevas tecnologías mientras intentan mantener a sus hijos seguros. Los padres deben mantenerse al día con los nuevos desarrollos de herramientas y comprender los riesgos potenciales que enfrentan sus hijos. También deben comprender las limitaciones de la prohibición de las redes sociales para proteger a los niños de daños.

Esto resalta la urgente necesidad de revisar la legislación propuesta por el deber de cuidado de Australia. Mientras que la Unión Europea y el Reino Unido lanzaron la legislación de Derechos de Cuidado de Digital de Cuidado en 2023, Australia ha estado en espera desde noviembre de 2024. Esta legislación haría que las empresas tecnológicas tengan en cuenta legislando que se ocupan de contenido nocivo, en la fuente, para proteger a todos.

Proporcionado por la conversación

Este artículo se vuelve a publicar de la conversación bajo una licencia Creative Commons. Lea el artículo original.La conversación

Citación: Google está implementando su chatbot Gemini Ai para niños menores de 13 años. Es un movimiento arriesgado (2025, 11 de mayo) recuperado el 11 de mayo de 2025 de https://techxplore.com/news/2025-05-google-gemini-ai-chatbot-kids.html

Este documento está sujeto a derechos de autor. Además de cualquier trato justo con el propósito de estudio o investigación privada, no se puede reproducir ninguna parte sin el permiso por escrito. El contenido se proporciona solo para fines de información.

Continue Reading

Trending