Technology

Geoffrey Hinton: "The AI problem is more urgent than climate change"
Nobel laureate in Physics Geoffrey Hinton, considered one of the "godfathers" of AI, warned at the Starmus Festival in Tenerife that the risk from artificial intelligence is more urgent than climate change, because its catastrophic consequences could arrive "in a few years". He said the probability that AI ends humanity lies between 1% and 99%, and that a figure of 10% or 20% is already "very worrying". He also urged Europe to build its own tech sector.
TechnologyInternationalLa VanguardiaOct 25AI
FOBO, the fear of becoming obsolete because of AI, hits senior professionals
FOBO (Fear of Becoming Obsolete) has become more visible with the rise of AI and particularly affects workers over 50. The World Economic Forum's Future of Jobs 2025 report estimates that 39% of core skills will have changed by 2030, and the OECD notes that workers aged 55 to 65 use fewer advanced digital tools and receive less training. Psychologists and trainers stress that experience remains key to supervising what AI produces.
TechnologyInternationalLa VanguardiaOct 25AI
Cambridge Analytica, ten years on: Meta cannot shake the ghosts of political interference
A decade after the Cambridge Analytica scandal, which collected data from 87 million Facebook users without consent, the debate over political manipulation on social media remains alive. Meta was ordered by a New Mexico court to pay a fine of about 22 billion euros for violating protection laws, on top of 4.5 billion already paid to the FTC. Meanwhile, Brazil's Superior Electoral Court is demanding that Meta explain an alleged favorable treatment of Flávio Bolsonaro over Lula da Silva.
TechnologyInternationalEL PAÍSOct 24AI
Tech stocks stumble and reopen the debate over who pays the AI bill
The slump of CoreWeave, an AI cloud computing firm backed by Nvidia, is dragging semiconductors into their worst streak since 2008 and reopening the debate over the sector's sustainability. CoreWeave is down 36% on the stock market over the past month and the yield demanded on its bonds has reached 16%, with liabilities of $50 billion. Semiconductors have fallen 19% over 12 sessions, while AI demand remains strong and external financing of the sector grows.
TechnologyInternationalCinco DíasOct 24AI
Bucaramanga installs 80 license plate reading cameras and more than 20 with facial recognition
Bucaramanga's city government has begun works to install 80 LPR cameras, which read vehicle license plates in real time, and more than 20 facial recognition devices, as part of a project of more than 100 cameras connected to the Metropolitan Police monitoring center. The announced investment is around 10 billion pesos, funded with the city's own resources, and includes new technology and the recovery of the existing network. According to Mayor Cristian Fernando Portilla Pérez, the equipment would be operating in December 2026.
TechnologyColombiaVanguardiaOct 24AI
Argentine neuroscientist creates an AI app that detects early Alzheimer's from minutes of speech
Argentine neuroscientist Adolfo García, director of the Cognitive Neuroscience Center at Universidad de San Andrés and of the Global Brain Health Institute in the United States, developed a method that uses AI to analyze brief samples of spontaneous speech to detect Alzheimer's and frontotemporal dementia before they become evident. The system measures rhythm, pauses and intonation in the audio and analyzes transcribed language. According to García, it exceeds 90% accuracy and is offered through the TELL app in more than 15 countries.
TechnologyInternationalClarin.comOct 24AI
Alexandr Wang, the 29-year-old floated as a possible successor to Zuckerberg at Meta
Alexandr Wang, 29, heads Meta's superintelligence labs and people close to the company's artificial intelligence area mention him as a possible successor to Mark Zuckerberg, according to the Financial Times. He took the role after a USD 14.3 billion deal that included a 49% stake in his company Scale AI. The launch of the Muse Wonder model drove a rise in Meta's shares, and the company announced a partnership with Apple to bring its open-weight models to Mac computers.
TechnologyInternationalInfobaeOct 24AI
Bogotá opens 500 spots in free English, AI and technology courses for people over 50
Bogotá's District Secretariat of Economic Development opened 500 spots in free courses for people over 50, as part of the Mayor's Office's Talento+50 strategy. The offering brings together three programs: Experta +50, with training in digital literacy, digital transformation, circular economy and finance; Experta Tech, focused on digital and technological skills with support from IBM, Oracle, AWS and Universidad Distrital; and YES Bogotá, with SENA A1 and A2 English courses. Registration is open until October 30, 2026.
TechnologyColombiaNoticias RCNOct 24AI
Pentagon adds Confucius Institutes and 80 linked entities to its national security threat list
U.S. Defense Secretary Pete Hegseth announced that the Pentagon added the Confucius Institutes and 80 linked entities to its list of organizations considered a threat to national security. According to Hegseth, these China-funded centers operated covertly at 30 universities in the country to steal intellectual property and influence students, and adopted other names to evade congressional restrictions. The Defense Department said it dismantled those organizations over the past two months.
TechnologyInternationalInfobaeOct 24AI
Uber, Google and Bolt prepare autonomous cars in Spain with no clear insurance model
Uber, Google and Bolt are already testing autonomous vehicles in Spanish cities, still with a human on board and under the DGT framework, and expect to operate autonomously in 2027. Insurers are working against the clock to define how to measure the risk of driverless fleets, with no claims history and with liability shifting to the manufacturer or the service operator, according to experts from Mapfre and Mutua Madrileña.
TechnologyInternationalABCOct 24AI
Mexico City Cyber Police warn of scam using view-once photos on WhatsApp
Mexico City's Cyber Police warned about a scam on WhatsApp involving view-once photographs sent from unknown numbers. Once the victim opens the image, cybercriminals claim it was sent by mistake or accuse the victim of accessing private content, then threaten complaints, legal action or the spread of information to demand payment. The SSC CDMX unit urged users not to open such files, not to transfer money or share data, and to report the number.
TechnologyInternationalUnoTVOct 24AI
Critically ill patient received AI-designed viruses to fight antibiotic-resistant bacteria
San Francisco-based company Radical Numerics announced that a patient received treatment with bacteriophages designed by artificial intelligence to eliminate the antibiotic-resistant bacteria infecting him. According to the company, it is the first time this type of therapy has been attempted, applied under compassionate use. The firm presented its Omnii-1 model and used it to create phage variants that eliminated the patient's isolate in vitro. No clinical results or the infection involved were published.
TechnologyInternationalBioBioChileOct 24AI
Musk expects digital intelligence to surpass human intelligence by 2031 and up to a billion humanoid robots
Elon Musk laid out a scenario of technological and economic transformations for the coming years. According to his remarks, by 2031 digital intelligence would surpass the sum of all human intelligence, and there could be between 100 million and a billion humanoid robots. He also estimated that within five to seven years the world economy would probably be twice its current size, driven by a phase of economic output doubling.
TechnologyInternationalRevista SemanaOct 24AI
PHP security flaw threatens much of the internet as millions of exploitation attempts are logged
Researchers discovered a vulnerability in PHP, the open-source language used by an estimated 70% of websites and web applications. The flaw, hidden for years, was identified by security firm Calif using OpenAI's GPT-5.6 Sol model and independently reported by Wiz. Cloudflare reported millions of exploitation attempts after the disclosure, and several government sites in Europe and Asia went offline as a precaution. As of Thursday afternoon, no successful attack had been confirmed.
TechnologyInternationalEl EspectadorOct 24AI
More than half of contact centers report heavier workloads for agents despite AI
A MaxContact survey included in its 2025/26 KPI report indicates that more than half of contact center managers say their agents' workload increased, compared with about 40% in 2024. The text attributes the phenomenon to automation absorbing routine tasks and leaving people with the most complex cases, and cites ArvatoConnect's 70/30 model and the "self-service paradox" described by Brad Cleveland.
TechnologyInternationalEl Ecosistema StartupOct 24AI
AI absorbs the simple tasks in contact centers and leaves agents the hardest cases
In some contact centers technology already resolves 70% of interactions and only 30% reaches a person, a split that, according to the text, leaves humans with the most complex and emotionally charged cases. ArvatoConnect targets that model with AI voice agents for routine work and advisors for disputes and vulnerable customers. A MaxContact report says more than half of contact center managers report higher workloads for their agents, versus around 40% in 2024.
TechnologyInternationalEl Ecosistema StartupOct 24AI
AI raises data center energy demand as industry responds with more efficient chips
The growth of artificial intelligence is increasing data centers' electricity demand, which according to the International Energy Agency consumed about 415 TWh in 2024, 1.5% of global consumption, and could exceed 945 TWh by 2030. Chipmakers are pursuing greater efficiency through advanced processes and dedicated AI units (NPUs), as well as 'on-device' AI. MediaTek's Dimensity 9600 Pro, built on 2 nanometers, reports up to 61% lower CPU power use in multi-core workloads.
TechnologyInternationalEl DiarioOct 24AI
NEO Consulting study: Prima, Habitat and Integra lead AFP visibility on ChatGPT and Gemini depending on the query type
A NEO Consulting study analyzed the visibility of Prima AFP, AFP Integra, AFP Habitat and Profuturo AFP on ChatGPT and Gemini using its AEO Tracker tool, with queries grouped into four categories. Prima AFP leads in informational queries, AFP Habitat in comparisons and AFP Integra in transfers. On withdrawals, Prima leads on Gemini and Integra on ChatGPT. No administrator leads every category and Profuturo records the lowest visibility in the group.
TechnologyInternationalGestiónOct 24AI
Icetex opens scholarships for Colombian startups: fully funded training in Switzerland
Icetex is offering scholarships under the Academia-Industry Training (AIT) program, which aims for Swiss and Colombian startups to innovate and connect researcher-entrepreneurs from both countries. Ten startup representatives, one per company, will be selected and can compete for COP 40,000,000 in support. Training will have a phase in Colombia from February 8 to 12, 2027, organized by Icetex, and another in Switzerland from March 15 to 19, 2027, led by the Swiss Leading House, with participation from the University of St. Gallen. The call closes on November 23, 2026.
TechnologyColombiaEl EspectadorOct 24AI
Global investment scam network exposed with more than 11,000 Spaniards affected
An investigation by IDEAL, based on documents leaked by OCCRP, reveals how an international cybercrime network running fake investments operates. Between 2022 and 2024 it collected sensitive data from more than 11,000 Spaniards, half of whom were scammed, and recorded victims in around twenty countries. The network lures retirees and vulnerable people through online ads leading to fake news articles made with artificial intelligence, and made more than 105 million euros in that period.
TechnologyInternationalIdealOct 24AI