El fin del trabajo

[muyuu]

Justicio: una IA que ofrece respuestas legales fundamentadas porque conoce toda la legislación estatal, autonómica y europea

Por @Alvy — 16 de Noviembre de 2024

Justicio: una IA que ofrece respuestas legales fundamentadas porque conoce toda la legislación estatal, autonómica y europea

Me encontré hace algunas semanas con Justicio, que supongo que no había visto antes porque no es mi especialidad ni mi área de interés, pero me pareció una idea genial. Es una plataforma gratuita con una IA que han alimentado con toda la legislación estatal, autonómica y europea. Preguntas cualquier tema legal y te lo explica. Se utiliza haciendo preguntas en lenguaje natural y puede responder con distintos niveles de complejidad.

Muy rápido no es, pero tras unos segundos desde que le lanzas una pregunta «¿Puedo instalar paneles solares en el tejado de mi casa si es un chalet independiente?» responde con toda la documentación necesaria (proyecto técnico, licencias…) o al menos con unas cuantas buenas pistas.

Tal y como explican en el «Acerca de» se trata de dar respuestas legales fundamentadas. Sus creadores han minimizado la posibilidad de error no permitiendo que genere documentos, entendiendo el contexto y almacenando sólo información que no esté relacionada con los temas legales.

Según sus creadores, domina todos los ámbitos relacionados con las leyes, así que da igual si se trata de una pregunta sobre normativas de edificación, un divorcio o un problema laboral con una excedencia. Se ha leído millones de documentos, y legislación desde el nivel local al europeo, BOEs incluidos. Para sacarle provecho hay que saber que tiene sus trucos y particularidades pero basta con leer la Ayuda para usarlo con mayor capacidad.

Hasta donde he podido probar funciona de maravilla, con respuestas muy claras y explicaciones razonadas que muchas veces van más allá de lo que has preguntado. Cita normas y leyes, entiendo que con precisión. Y cuando «bajas el nivel» (puede contestar con nivel de abogado, adulto, joven, niño) se adapta bastante bien. Incluso en cuestiones más atípicas como «¿Se puede caminar desnudo por calle?» o «¿Puedo sacrificar un toro en una plaza de toros?» lo que explica suena bien.

Además de todo esto su Manifiesto aboga por el código abierto. Tanto el diseño como la herramienta en sí se basan en principios de universalidad, independencia, transparencia y uso responsable… No sólo eso, sino que fomenta la colaboración con otros usuarios y busca siempre estar a la última en tecnología. Los principios más claros, abiertos y transparentes que se pueden esperar de un proyecto y que debería tener la propia administración de justicia en el país, en vez de trabajar con webs lamentablemente confusas, cerradas y buscadores inútiles, vamos.

El número de consultas está limitado a unas pocas diarias, según el «principio de responsabilidad de uso» que proponen, básicamente para poder compartir los recursos (limitados) entre muchas personas con cierta calidad. Aun así pueden ayudarte a instalar tu propia plataforma (al ser código libre) o incluso desbloquear esos límites, dependiendo de los casos.

Saludos.

A Justicio se le ha olvidado mencionar que es perfectamente legal irse a Andorra y pasar más de medio año fuera de España, declarando por completo allí.

[tomasjos]

Justicio: una IA que ofrece respuestas legales fundamentadas porque conoce toda la legislación estatal, autonómica y europea

Por @Alvy — 16 de Noviembre de 2024

Yo la he probado. No sirve de nada, no conoce la legislación autonómica asturiana de educaciòn Justicio: una IA que ofrece respuestas legales fundamentadas porque conoce toda la legislación estatal, autonómica y europea

Me encontré hace algunas semanas con Justicio, que supongo que no había visto antes porque no es mi especialidad ni mi área de interés, pero me pareció una idea genial. Es una plataforma gratuita con una IA que han alimentado con toda la legislación estatal, autonómica y europea. Preguntas cualquier tema legal y te lo explica. Se utiliza haciendo preguntas en lenguaje natural y puede responder con distintos niveles de complejidad.

Muy rápido no es, pero tras unos segundos desde que le lanzas una pregunta «¿Puedo instalar paneles solares en el tejado de mi casa si es un chalet independiente?» responde con toda la documentación necesaria (proyecto técnico, licencias…) o al menos con unas cuantas buenas pistas.

Tal y como explican en el «Acerca de» se trata de dar respuestas legales fundamentadas. Sus creadores han minimizado la posibilidad de error no permitiendo que genere documentos, entendiendo el contexto y almacenando sólo información que no esté relacionada con los temas legales.

Según sus creadores, domina todos los ámbitos relacionados con las leyes, así que da igual si se trata de una pregunta sobre normativas de edificación, un divorcio o un problema laboral con una excedencia. Se ha leído millones de documentos, y legislación desde el nivel local al europeo, BOEs incluidos. Para sacarle provecho hay que saber que tiene sus trucos y particularidades pero basta con leer la Ayuda para usarlo con mayor capacidad.

Hasta donde he podido probar funciona de maravilla, con respuestas muy claras y explicaciones razonadas que muchas veces van más allá de lo que has preguntado. Cita normas y leyes, entiendo que con precisión. Y cuando «bajas el nivel» (puede contestar con nivel de abogado, adulto, joven, niño) se adapta bastante bien. Incluso en cuestiones más atípicas como «¿Se puede caminar desnudo por calle?» o «¿Puedo sacrificar un toro en una plaza de toros?» lo que explica suena bien.

Además de todo esto su Manifiesto aboga por el código abierto. Tanto el diseño como la herramienta en sí se basan en principios de universalidad, independencia, transparencia y uso responsable… No sólo eso, sino que fomenta la colaboración con otros usuarios y busca siempre estar a la última en tecnología. Los principios más claros, abiertos y transparentes que se pueden esperar de un proyecto y que debería tener la propia administración de justicia en el país, en vez de trabajar con webs lamentablemente confusas, cerradas y buscadores inútiles, vamos.

El número de consultas está limitado a unas pocas diarias, según el «principio de responsabilidad de uso» que proponen, básicamente para poder compartir los recursos (limitados) entre muchas personas con cierta calidad. Aun así pueden ayudarte a instalar tu propia plataforma (al ser código libre) o incluso desbloquear esos límites, dependiendo de los casos.

Saludos.

A Justicio se le ha olvidado mencionar que es perfectamente legal irse a Andorra y pasar más de medio año fuera de España, declarando por completo allí.

[/td][/tr][/table]

[Cadavre Exquis]

Waymo Robotaxis Are Coming To Tokyo In 2025
Posted by BeauHD on Tuesday December 17, 2024 @05:00AM from the what-to-expect dept.

Waymo will begin testing its autonomous vehicles in Tokyo in early 2025, marking its first deployment outside the U.S. TechCrunch reports:

The move to Japan is part of Waymo's "road trips," a development program that involves bringing its technology to a variety of cities and testing it -- with each city having different challenges. In Tokyo, the Waymo robotaxis will face left-hand driving and a dense urban environment. [...] Waymo said it will partner with taxi-hailing app GO and taxi company Nihon Kotsu as part of its Japanese "road trip." Nihon Kotsu will oversee the management and servicing of the Waymo vehicles, according to the company.

Initially, Nihon Kotsu drivers will operate the vehicles manually to map key areas of the Japanese capital, including Minato, Shinjuku, Shibuya, Chiyoda, Chuo, Shinagawa, and Koto. Waymo said it is working with Nihon Kotsu's team to train its employees how to operate Waymo's self-driving Jaguar I-Pace vehicles.

Saludos.

[pollo]

Waymo Robotaxis Are Coming To Tokyo In 2025
Posted by BeauHD on Tuesday December 17, 2024 @05:00AM from the what-to-expect dept.

Waymo will begin testing its autonomous vehicles in Tokyo in early 2025, marking its first deployment outside the U.S. TechCrunch reports:

The move to Japan is part of Waymo's "road trips," a development program that involves bringing its technology to a variety of cities and testing it -- with each city having different challenges. In Tokyo, the Waymo robotaxis will face left-hand driving and a dense urban environment. [...] Waymo said it will partner with taxi-hailing app GO and taxi company Nihon Kotsu as part of its Japanese "road trip." Nihon Kotsu will oversee the management and servicing of the Waymo vehicles, according to the company.

Initially, Nihon Kotsu drivers will operate the vehicles manually to map key areas of the Japanese capital, including Minato, Shinjuku, Shibuya, Chiyoda, Chuo, Shinagawa, and Koto. Waymo said it is working with Nihon Kotsu's team to train its employees how to operate Waymo's self-driving Jaguar I-Pace vehicles.

Saludos.

Van a tener que contratar un ejército de pringaos para toda la flota.

No conozco Tokyo, pero entiendo que por cultura de respeto a las normas será uno de los entornos menos hostiles para los robotaxis estos, que por cierto, siguen necesitando que les saquen por control remoto de las meteduras de pata bastante frecuentes en las que caen.

Lo último que yo he leído (quitando el publirreportaje -con mamada incluída en el titular- de Wired, en su línea habitual de que todo es maravillosísimo) es que la gente les tenía bastante tirria y no sólo por las cagadas de las máquinas, sino contra la empresa en sí.

[Cadavre Exquis]

https://x.com/DotCSV/status/1869491144767455469

Saludos.

[puede ser]

Desde que me tiré por la ventana llevo dos segundos sin estrellarme... LOL, el vuelo sin motor ya es una realidad 

[Cadavre Exquis]

Waymo's Driverless Cars Are Apparently an Insurance Company's Dream
Posted by BeauHD on Thursday December 19, 2024 @07:50PM from the would-you-look-at-that dept.

A study by reinsurer Swiss Re found that Waymo's autonomous vehicles have demonstrated significantly fewer property damage and bodily injury claims compared to human-driven cars, with reductions of up to 92% in some metrics. Engadget reports:

Swiss Re analyzed liability claims from collisions covering 25.3 million miles driven by Waymo's autonomous cars. The study also compared Waymo's liability claims to human driver baselines based on data from over 500,000 claims and over 200 billion driving miles. The results found that Waymo Driver "demonstrated better safety performance when compared to human-driver vehicles." The study found cars operated by Alphabet's Waymo Driver resulted in 88 percent fewer property damage claims and 92 percent fewer bodily injury claims.

Swiss Re also invented a new metric to compare Waymo Driver against only newer vehicles with advanced safety tech, like driver assistance, automated emergency braking and blind spot warning systems, instead of against the whole corpus of those 200 billion driving miles. In this comparison, Waymo still came out ahead with an 86 percent reduction in property damage claims and a 90 percent reduction on bodily damage claims.

Saludos.

[puede ser]

Waymo's Driverless Cars Are Apparently an Insurance Company's Dream
Posted by BeauHD on Thursday December 19, 2024 @07:50PM from the would-you-look-at-that dept.

A study by reinsurer Swiss Re found that Waymo's autonomous vehicles have demonstrated significantly fewer property damage and bodily injury claims compared to human-driven cars, with reductions of up to 92% in some metrics. Engadget reports:

Swiss Re analyzed liability claims from collisions covering 25.3 million miles driven by Waymo's autonomous cars. The study also compared Waymo's liability claims to human driver baselines based on data from over 500,000 claims and over 200 billion driving miles. The results found that Waymo Driver "demonstrated better safety performance when compared to human-driver vehicles." The study found cars operated by Alphabet's Waymo Driver resulted in 88 percent fewer property damage claims and 92 percent fewer bodily injury claims.

Swiss Re also invented a new metric to compare Waymo Driver against only newer vehicles with advanced safety tech, like driver assistance, automated emergency braking and blind spot warning systems, instead of against the whole corpus of those 200 billion driving miles. In this comparison, Waymo still came out ahead with an 86 percent reduction in property damage claims and a 90 percent reduction on bodily damage claims.

Saludos.

Esto está mejor... Pero, ¿la comparación es entre el mismo tipo de desplazamientos?

Y luego está otro tema: hay conductores humanos en ciertos momentos con los que nadie con dos dedos de frente se subiría al coche?, y esos también suman a la estadística de accidentes humanos.

Una comparación interesante de siniestralidad sería entre Waymo y taxistas que hagan rutas urbanas semejantes.

[Cadavre Exquis]

https://cincodias.elpais.com/opinion/2024-12-24/la-fusion-honda-nissan-avanza-con-demasiada-lentitud.html

Saludos.

[Cadavre Exquis]

Man Trapped in Circling Waymo on Way to Airport
Posted by EditorDavid on Sunday January 05, 2025 @08:49PM from the I-love-L.A. dept.

It "felt like a Disneyland ride," reports CBS News. A man took a Waymo takes to the airport — only to discover the car "wouldn't stop driving around a parking lot in circles." And because the car was in motion, he also couldn't get out.

Still stuck in the car, Michael Johns — a tech-industry worker — then phoned Waymo for help. ("Has this been hacked? What's going on? I feel like I'm in the movies. Is somebody playing a joke on me?") But he also filmed the incident...

"Why is this thing going in a circle? I'm getting dizzy," Johns said in a video posted on social media that has since gone viral, garnering more than two million views and interactions....

The Waymo representative was finally able to get the car under control after a few minutes, allowing him to get to the airport just in time to catch his flight back to LA. He says that the lack of empathy from the representative who attempted to help him, on top of the point that he's unsure if he was talking to a human or AI, are major concerns. "Where's the empathy? Where's the human connection to this?" Johns said while speaking with CBS News Los Angeles. "It's just, again, a case of today's digital world. A half-baked product and nobody meeting the customer, the consumers, in the middle."

Johns, who ironically works in the tech industry himself, says he would love to see services like Waymo succeed, but he has no plans to hop in for a ride until he's sure that the kinks have been fixed. In the meantime, he's still waiting for someone from Waymo to contact him in regards to his concerns, which hasn't yet happened despite how much attention his video has attracted since last week.

"My Monday was fine till i got into one of Waymo 's 'humanless' cars," he posted on LinkedIn. "I get in, buckle up ( safety first) and the saga begins.... [T]he car just went around in circles, eight circles at that..."

A Waymo spokesperson admitted they'd added about five minutes to his travel time, but then "said the software glitch had since been resolved," reports the Los Angeles Times, "and that Johns was not charged for the ride."

One final irony? According to his LinkedIn profile, Johns is a CES Innovations Awards judge.

Saludos.

[pollo]

Man Trapped in Circling Waymo on Way to Airport
Posted by EditorDavid on Sunday January 05, 2025 @08:49PM from the I-love-L.A. dept.

It "felt like a Disneyland ride," reports CBS News. A man took a Waymo takes to the airport — only to discover the car "wouldn't stop driving around a parking lot in circles." And because the car was in motion, he also couldn't get out.

Still stuck in the car, Michael Johns — a tech-industry worker — then phoned Waymo for help. ("Has this been hacked? What's going on? I feel like I'm in the movies. Is somebody playing a joke on me?") But he also filmed the incident...

"Why is this thing going in a circle? I'm getting dizzy," Johns said in a video posted on social media that has since gone viral, garnering more than two million views and interactions....

The Waymo representative was finally able to get the car under control after a few minutes, allowing him to get to the airport just in time to catch his flight back to LA. He says that the lack of empathy from the representative who attempted to help him, on top of the point that he's unsure if he was talking to a human or AI, are major concerns. "Where's the empathy? Where's the human connection to this?" Johns said while speaking with CBS News Los Angeles. "It's just, again, a case of today's digital world. A half-baked product and nobody meeting the customer, the consumers, in the middle."

Johns, who ironically works in the tech industry himself, says he would love to see services like Waymo succeed, but he has no plans to hop in for a ride until he's sure that the kinks have been fixed. In the meantime, he's still waiting for someone from Waymo to contact him in regards to his concerns, which hasn't yet happened despite how much attention his video has attracted since last week.

"My Monday was fine till i got into one of Waymo 's 'humanless' cars," he posted on LinkedIn. "I get in, buckle up ( safety first) and the saga begins.... [T]he car just went around in circles, eight circles at that..."

A Waymo spokesperson admitted they'd added about five minutes to his travel time, but then "said the software glitch had since been resolved," reports the Los Angeles Times, "and that Johns was not charged for the ride."

One final irony? According to his LinkedIn profile, Johns is a CES Innovations Awards judge.

Saludos.

Alguien lo describió muy bien en los comentarios de Slashdot hace unos días: estos modelos conocen de todo pero no son capaces de comprender nada.

[Cadavre Exquis]

StackOverflow Usage Plummets as AI Chatbots Rise
Posted by msmash on Friday January 10, 2025 @12:34PM from the tough-luck dept.

Developer Q&A platform StackOverflow appears to be facing an existential crisis as volume of new questions on the site has plunged 75% from the 2017 peak and 60% year-on-year in December 2024, according to StackExchange Data Explorer figures.

The decline accelerated after ChatGPT's launch in November 2022, with questions falling 76% since then. Despite banning AI-generated answers two years ago, StackOverflow has embraced AI partnerships, striking deals with Google, OpenAI and GitHub.

Saludos.

[pollo]

StackOverflow Usage Plummets as AI Chatbots Rise
Posted by msmash on Friday January 10, 2025 @12:34PM from the tough-luck dept.

Developer Q&A platform StackOverflow appears to be facing an existential crisis as volume of new questions on the site has plunged 75% from the 2017 peak and 60% year-on-year in December 2024, according to StackExchange Data Explorer figures.

The decline accelerated after ChatGPT's launch in November 2022, with questions falling 76% since then. Despite banning AI-generated answers two years ago, StackOverflow has embraced AI partnerships, striking deals with Google, OpenAI and GitHub.

Saludos.

Es interesante cómo se quiere cambiar sin pestañear el obtener respuestas 100% ciertas de gente verdaderamente experta por resúmenes posiblemente falsos de una máquina que no sabe lo que hace.

El futuro va a ser interesante. Pero en el mal sentido, que no se me malinterprete. Dentro de 15-20 años veremos qué tipo de trabajadores o creadores tendremos. El proceso ya empezó hace tiempo, esto sólo hará que el camino a la imbecilidad absoluta sea mucho más eficiente.

Cada vez hay menos y menos incentivos para desarrollar o entrenar habilidades o el intelecto. Hasta que las sociedades acaben aplastadas por su propio peso idiocrático, o por otras sociedades que no hayan caído en la trampa.

La era de la imbecilidad ya se ha establecido. Todos, hasta los más inteligentes, claman por la utopía repleta de imbecilidad, en la que no haya que hacer ni pensar nada. Y a esto se le llama "progreso", "avance" y "futuro".

Esto es análogo a cómo funcionan las drogas más duras. Una vez que el cerebro se acostumbra a obtener estímulo por el medio exógeno, no buscará otra cosa, aunque sea perjudicial o en los casos más extremos se autodestruya en el proceso.

https://gizmodo.com/the-curious-case-of-a-woman-addicted-to-her-brain-impla-5402584

Pues nada, de cabeza sin cuestionar nada. A aprender otra dura lección.

[Cadavre Exquis]

https://x.com/sninobecerra/status/1877750616354627962

Saludos.

[senslev]

https://techcrunch.com/2025/01/08/elon-musk-agrees-that-weve-exhausted-ai-training-data/

Elon Musk concurs with other AI experts that there’s little real-world data left to train AI models on.

“We’ve now exhausted basically the cumulative sum of human knowledge … in AI training,” Musk said during a livestreamed conversation with Stagwell chairman Mark Penn on X late Wednesday. “That happened basically last year.”

Musk, who owns AI company xAI, echoed themes former OpenAI chief scientist Ilya Sutskever touched on at NeurIPS, the machine learning conference, during an address in December. Sutskever, who said the AI industry had reached what he called “peak data,” predicted a lack of training data will force a shift away from the way models are developed today.

Indeed, Musk suggested that synthetic data — data generated by AI models themselves — is the path forward. “The only way to supplement [real-world data] is with synthetic data, where the AI creates [training data],” he said. “With synthetic data … [AI] will sort of grade itself and go through this process of self-learning.”

Other companies, including tech giants like Microsoft, Meta, OpenAI, and Anthropic, are already using synthetic data to train flagship AI models. Gartner estimates 60% of the data used for AI and an­a­lyt­ics projects in 2024 were syn­thet­i­cally gen­er­ated.

Microsoft’s Phi-4, which was open sourced early Wednesday, was trained on synthetic data alongside real-world data. So were Google’s Gemma models. Anthropic used some synthetic data to develop one of its most performant systems, Claude 3.5 Sonnet. And Meta fine-tuned its most recent Llama series of models using AI-generated data.

Training on synthetic data has other advantages, like cost savings. AI startup Writer claims its Palmyra X 004 model, which was developed using almost entirely synthetic sources, cost just $700,000 to develop — compared to estimates of $4.6 million for a comparably sized OpenAI model.

But there as disadvantages as well. Some research suggests that synthetic data can lead to model collapse, where a model becomes less “creative” — and more biased — in its outputs, eventually seriously compromising its functionality. Because models create synthetic data, if the data used to train these models has biases and limitations, their outputs will be similarly tainted.