Rogue AI goes Admin-mode, installs updates that the PC could never reboot from:
A recent incident highlights the importance of using AI carefully and responsibly ...
neowin.net/news/rogue-ai-goes-admin-mode-installs-updates-that-the-pc-could-never-reboot-from/
Story Details
| Last Updated | Oct 02 |
| Total Articles | 2 |
AI agent promotes itself to sysadmin, trashes boot sequence: Fun experiment, but yeah, don't pipe an LLM raw into /bin/bash
theregister.com/2024/10/02/ai_agent_trashes_pc/
More articles on this subject
Previous
AI agent promotes itself to sysadmin, trashes boot sequence: Fun experiment, but yeah, don't pipe an LLM raw into /bin/bash
theregister.com/2024/10/02/ai_agent_trashes_pc/
Tesla's Optimus robots revealed to be human-controlled at "We, Robot" event
techspot.com/news/105119-tesla-optimus-robots-revealed-human-controlled-robot-event.html
Cybercab event: Tesla's Optimus robots are more of a show than AI - | Golem.de
golem.de/news/cybercab-event-teslas-optimus-roboter-sind-mehr-show-als-ki-2410-189786.html
Meta gets roasted on Threads for suggesting you should fake Northern Lights photos
techradar.com/computing/artificial-intelligence/missed-the-northern-lights-meta-says-you-should-just-fake-photos-with-its-ai-instead
Super Realistic AI Scam Targets Billions Of Gmail Users, Here’s How It Works: A new warning has been issued by Microsoft solutions consultant Sam Mitrovic ...
hothardware.com/news/realistic-ai-scam-targets-gmail-users
The best use for those latest manycore chips? AI, say server vendors: PC makers might not be able to sell the idea - big iron has a better chance
theregister.com/2024/10/14/manycore_chips_ai_servers/
How Tesla's robots delivered drinks and smiles, but were 'not machines': After Elon Musk launched the humanoid robot, Optimus, an attendee at the California ...
business-standard.com/world-news/how-tesla-s-robots-delivered-drinks-and-smiles-but-were-not-machines-124101400457_1.html
Casio brings $400 AI-powered emotional support pet to market: Cutest robot ever?
mashable.com/article/ai-pet-companion-moflin-casio
Sony developers share how machine learning can improve QA: During a talk given at the recent CEDEC event in Yokohama, Japan, development leads within Sony ...
gamesindustry.biz/sony-developers-share-how-machine-learning-can-improve-qa
I Launched the AI Safety Clock. Here’s What It Tells Us About Existential Risks: The Clock’s current reading—29 minutes to midnight—shows just how close we are ...
time.com/7086139/ai-safety-clock-existential-risks/
Scientists who built ‘foundation’ for AI awarded Nobel Prize: The Nobel Prize in physics was awarded to two scientists credited with building the “foundation” ...
theverge.com/2024/10/8/24265066/ai-nobel-prize-winners-physics-machine-learning-regret
The AI future from a 70-year-old Roald Dahl story is here – are you afraid?: ‘The Great Automatic Grammatizator’ warned of a fantastic machine that could churn ...
scroll.in/article/1074225/the-ai-future-from-a-70-year-old-roald-dahl-story-is-here-are-you-afraid
Elon Musk’s Optimus bot stole the show at Tesla’s robotaxi unveil—but the AI was all smoke and mirrors
fortune.com/2024/10/13/elon-musk-tesla-optimus-robot-tele-operated-robotaxi/
Paid articlePaywall
Quick Glance: Tesla's New Robotaxi Design Causes Confusion
- Tesla unveiled a sporty, two-seater robotaxi that perplexed investors and experts.
- Criticism of the design and lack of details led to a 9% drop in Tesla's stock.
- Experts argue that robotaxis should offer more space, and the two-door design poses challenges for older passengers.
- Musk aims to make robotaxis cheaper than public transit but provided no clear production or approval timelines.
Asus ProArt P16 (H7606) Review: Powered by AMD's latest mobile AI processor and hearty Nvidia graphics, the Asus ProArt P16 is an unrivaled content-creator ...
pcmag.com/reviews/asus-proart-p16-h7606
t3n.de 3h
Power for AI data centers: Google relies on nuclear power plant made of new reactors
t3n.de/news/strom-fuer-ki-rechenzentren-google-setzt-auf-atomkraftwerk-aus-neuartigen-reaktoren-1651542/
Quick Glance: Google Turns to Nuclear Power for Data Centers
- Google plans to source energy from small reactors by Kairos Power starting in 2030 to meet the high electricity demands of Artificial Intelligence.
- The goal is to achieve an annual output of 500 megawatts by 2035, with uncertainties regarding the integration of the reactors into the power grid.
- Kairos' compact reactors use molten fluoride salts for cooling, making them safer than traditional designs.
- Despite climate commitments, Google's CO2 emissions rose by 13 percent last year, driven by the significant energy consumption of its data centers.
BBC 9h
Google turns to nuclear to power AI data centres: The tech giant says it will use energy from small reactors to power its use of artificial intelligence.
bbc.com/news/articles/c748gn94k95o
Quick Glance: Google Secures Nuclear Energy Deal with Kairos Power
- Google has signed a deal to obtain electricity from small nuclear reactors designed to support AI.
- The first of these modular reactors is expected to be operational by the end of the decade, contributing to a total of 500 megawatts.
- Nuclear power is viewed as crucial for meeting the growing energy demands of AI while reducing reliance on renewable sources.
- Despite safety concerns and high costs, tech companies are optimistic about nuclear energy's potential to evolve and meet future electricity needs.
Britain opens floodgates to US datacenter investment: Who needs climate goals and planning permission anyway?
theregister.com/2024/10/15/uk_datacenter_investment/
Apple Reportedly Developing Smart Glasses to Rival Meta's Ray-Bans: But a 2027 timeline could give Meta the advantage for years.
cnet.com/tech/mobile/apple-reportedly-developing-smart-glasses-to-rival-metas-ray-bans/
NVDA closes on record high, Nvidia could top Apple as world’s most valuable company: Here’s what’s driving the latest rally
fastcompany.com/91209304/nvidia-stock-nvda-closes-record-high-most-valuable-company-whats-driving-latest-rally
Quick Glance: Nvidia Considers Sockets for AI GPUs, Not GeForce
- Nvidia may switch its upcoming B300 Blackwell AI chips from directly mounted processors to a socketed design to simplify maintenance and servicing.
- Trendforce notes that this change would benefit Nvidia's AI hardware manufacturers by potentially reducing the need for surface-mount machinery.
- However, there will likely be no socketed GPUs for consumers in the near future, as RAM is integrated with the processing chiplets.
- The complexity and cost of a socket-based GPU system are too high due to varying memory architectures and the lack of standards among AMD, Intel, and Nvidia.
From Bias to Balance: AI’s Role in Workplace Equity: The latest episode of Mint’s AI series highlighted AI’s potential to create equitable workplaces ...
livemint.com/ai/from-bias-to-balance-ais-role-in-workplace-equity-11728930933820.html
Head of Saudi tech institute pledges to limit China AI collaboration: Kaust’s new chief promises to stop any deals that jeopardise access to US-made chips
ft.com/content/1652e551-dac7-40a5-9621-7453d13af613
Stephen Ibaraki, international tech consultant: ‘We need to prepare for what will happen in five years’
english.elpais.com/technology/2024-10-14/stephen-ibaraki-international-tech-consultant-we-need-to-prepare-for-what-will-happen-in-five-years.html
South Korea’s Cooling Memory Chip Prices Hint at a Peaking Out: Price gains for South Korea’s memory-chip exports slowed last month in a potential early warning ...
bloomberg.com/news/articles/2024-10-15/south-korea-s-cooling-memory-chip-prices-hint-at-a-peaking-out
Exclusive | New York Times to Bezos-Backed AI Startup: Stop Using Our Stuff: Publisher demands that Perplexity stop accessing its content. Startup isn’t ...
on.wsj.com/40bPHOW
Chinese think tank advises data centers to stick with Nvidia AI GPUs — homegrown Chinese GPUs suffer from 'high costs' and 'complex engineering'
tomshardware.com/pc-components/gpus/chinese-think-tank-advises-data-centers-to-stick-with-nvidia-ai-gpus-homegrown-chinese-gpus-suffer-from-high-costs-and-complex-engineering
Adobe Launches AI Video Tool, Rivals OpenAI And Meta : Round-up of top and latest news from the world of technology: - Adobe Launches AI Video Tool ...
wionews.com/videos/adobe-launches-ai-video-tool-rivals-openai-and-meta-767494
Think about it: AI pervades science Nobels : Think about it: AI
pervades science Nobels
deccanherald.com/opinion/editorial/think-about-it-ai-pervades-science-nobels-3232237
t3n.de 23h
Apple study: LLM-based AI models cannot think logically: Minor changes in the wording of requests to AI bots can lead to significantly different answers
t3n.de/news/apple-studie-llm-basierte-ki-modelle-koennen-nicht-logisch-denken-1651431/
TSMC is building new chip-making plants in Europe to get away from U.S. - China tensions
neowin.net/news/tsmc-is-building-new-chip-making-plants-in-europe-to-get-away-from-us---china-tensions/
Why data centres are a bottleneck for the AI boom: podcast: Artificial intelligence tools like ChatGPT depend on chips and energy. But a $1 trln rush to build ...
reuters.com/breakingviews/why-data-centres-are-bottleneck-ai-boom-podcast-2024-10-15/
Why data centres are a bottleneck for the AI boom: podcast: Artificial intelligence tools like ChatGPT depend on chips and energy. But a $1 trln rush to build ...
reuters.com/breakingviews/why-data-centres-are-bottleneck-ai-boom-podcast-2024-10-15/
Meta's AI chief Yann LeCun calls AI apocalypse fears "complete B.S."
techspot.com/news/105123-meta-ai-chief-yann-lecun-calls-ai-apocalypse.html
Adobe launches massive training program to fight digital skills gap: Adobe is aiming to train 30 million people of all ages worldwide by 2030 to help improve ...
usatoday.com/story/tech/columnists/2024/10/14/adobe-digital-marketing-initiative-skills-new-learners/75672266007/
US mulls capping Nvidia AI chips exports to some countries, Bloomberg News reports
reuters.com/technology/us-mulls-capping-nvidia-ai-chips-exports-some-countries-bloomberg-news-reports-2024-10-15/
Human foot found on Everest may hold key to one of mountaineering’s greatest mysteries
cnn.com/2024/10/11/science/sandy-irvine-boot-discovery-mount-everest-intl-scli/
Quick Glance: Everest Foot Discovery Unveils Mountain Mystery
- A National Geographic documentary team discovered a boot and sock protruding from a melting glacier on Mount Everest in September.
- The sock bore the name label “A.C. Irvine,” suggesting it belonged to British climber Andrew “Sandy” Irvine, who disappeared in 1924.
- This finding may provide critical clues about Irvine’s fate, as neither his body nor the camera they carried has been recovered.
- Irvine's family was informed of the discovery, prompting emotional reactions from his great-niece Julie Summers, who has long been connected to the story.
‘Freemium model, Salesforce base key to Slack’s India growth’: Rahul Sharma, Slack
financialexpress.com/business/industry-freemium-model-salesforce-base-key-to-slacks-india-growth-rahul-sharma-slack-india-country-manager-3638488/
New AI Scam Targets Gmail Users with Fake Account Recovery Requests : As hackers continue to evolve their tactics, Gmail users are facing increasingly ...
wionews.com/videos/new-ai-scam-targets-gmail-users-with-fake-account-recovery-requests-767337
AI's thirst for power keeps coal fires burning bright: So much for reducing our reliance on fossil fuels
theregister.com/2024/10/14/ai_datacenters_coal/
Westminster’s reliance on Elon Musk’s X is ‘totally wrong’, says Labour MP: Exclusive: Josh Simons, close ally of Starmer, suggests Musk manipulates platform’s ...
theguardian.com/technology/2024/oct/14/westminster-elon-musk-x-labour-mp-josh-simons
Unlocking AI-Powered Learning | by Justina Nixon-Saintil - Project Syndicate: Justina Nixon-Saintil foresees the technology enhancing student learning and ...
project-syndicate.org/commentary/ai-will-transform-education-enhance-student-learning-by-justina-nixon-saintil-2024-10
Adobe’s AI video model is here, and it’s already inside Premiere Pro: Adobe’s Firefly Video AI generation model is launching today, powering tools inside ...
theverge.com/2024/10/14/24268695/adobe-ai-video-generation-firefly-model-premiere-pro
Q3 earnings outlook appears rosy for Naver, gloomy for Kakao : The outlook for third-quarter earnings of Korea's leading internet service providers ...
koreatimes.co.kr/www/tech/2024/10/133_384225.html
Invisible text that AI chatbots understand and humans can’t? Yep, it’s a thing.: A quirk in the Unicode standard harbors an ideal steganographic code channel.
arstechnica.com/security/2024/10/ai-chatbots-can-read-and-write-invisible-text-creating-an-ideal-covert-channel/
Adobe unveils AI video generator trained on licensed content: New text-to-video tool focuses on video pros, made with content owner permission.
arstechnica.com/ai/2024/10/adobe-unveils-ai-video-generator-trained-on-licensed-content/
Why Dubai-based VCs are looking to invest in European climate tech: Rising temperatures and unpredictable weather patterns have become a pressing concern within ...
thenextweb.com/news/dubai-vcs-invest-european-climate-tech
Wccftech 15h
TSMC 3nm Set To Witness Massive Adoption From AI Tech Giants; NVIDIA Rubin, AMD Instinct MI355X & I…
wccftech.com/tsmc-3nm-witness-massive-adoption-ai-tech-giants-nvidia-rubin-amd-mi355x-intel-falcon-shores/
Apple study reveals major AI flaw in OpenAI, Google, and Meta LLMs: Their reasoning skills may not be as advanced as they seem.
mashable.com/article/apple-study-reveals-major-ai-flaw-openai-google-meta-llms
Quick Glance: Apple Study Reveals Flaws in AI Reasoning
- A new study by Apple engineers indicates that large language models (LLMs) exhibit extreme fragility and unreliability in logical reasoning.
- The researchers found that accuracy dropped significantly with minor alterations to test questions, supporting the notion that LLMs do not engage in genuine logical reasoning.
- Particularly, adding irrelevant information resulted in catastrophic performance declines, as models attempted to mimic patterns from their training data.
- The findings suggest deeper issues in the reasoning processes of these models that cannot be remedied through fine-tuning.
America v China: who controls Asia’s internet?: Amid an explosive data and AI boom the superpower contest hots up
economist.com/asia/2024/10/08/america-v-china-who-controls-asias-internet
Paid articlePaywall
Japan tech firms tap AI to help workers handle abusive customers - The Mainichi: TOKYO (Kyodo) -- Japanese tech companies NTT Communications Corp ...
mainichi.jp/english/articles/20241014/p2g/00m/0bu/018000c
Quick Glance: AI Systems Aid Call Center Workers Facing Abusive Customers
- Japanese tech firms NTT Communications and SoftBank are developing AI systems to assist call center workers in managing abusive customers.
- NTT Communications has created a system that monitors customer-employee exchanges and displays appropriate response examples on operators' screens.
- These systems aim to reduce employees' psychological stress and facilitate quicker responses to customer inquiries.
- The rise of customer abuse, referred to as 'kasu-hara,' has led to increased mental health issues among victims, prompting many companies to establish guidelines for addressing such behavior.
AI isn't likely to replace humans at work, says report—but 5 occupations may soon feel its impact more
cnbc.com/2024/10/14/ai-isnt-ready-to-replace-human-jobs-5-professions-may-feel-its-impact-more.html
We Regret to Inform You That the Perpetually Sold-Out Ninja Slushi Lives Up to the Hype
nytimes.com/wirecutter/reviews/ninja-slushi-review/
How Crunchbase’s AI Integration Helps Decision Making: In his first year at the company, Chief Product Officer Megh Gautam has added AI so users from CEOs to ...
forbes.com/sites/meganpoinski/2024/10/14/how-crunchbases-ai-integration-helps-decision-making/
AI amplifies systemic risk to financial sector, says India's Reserve Bank boss: Who also worries misinformation on social media could threaten liquidity
theregister.com/2024/10/15/india_rbi_ai_risks/
Concourse is building AI to automate financial tasks | TechCrunch: Concourse, a company building AI-powered automation tools for finance, has emerged from ...
tcrn.ch/3BMAMAU
Microsoft may have an audio-to-image generator in the works, new patent shows: Your meetings could soon be enhanced with live image generation.
zdnet.com/article/microsoft-may-have-an-audio-to-image-generator-in-the-works-new-patent-shows/
Copyright and AI training: | "In fact, it's almost a paradoxical situation": Is it legal for AI companies to use content from creatives for training without ...
heise.de/hintergrund/Urheberrecht-und-KI-Training-In-der-Tat-ein-fast-schon-paradoxer-Zustand-9979372.html
Apple's future products could include affordable smart glasses and IoT devices
techspot.com/news/105126-apple-future-products-could-include-affordable-smart-glasses.html
Indonesia orders Apple, Google to take down Chinese bargain app Temu: Plus: Infosys stops sending job offer emails; Singtel outage; Australia to require ...
theregister.com/2024/10/13/asia_tech_news_roundup/
Digital attacks: Secret services warn parliament of "hurricane": The hearing of the secret service chiefs in the Bundestag is being used to assess the situation ...
heise.de/news/Digitale-Angriffe-Geheimdienste-waren-Parlament-vor-Hurrikan-9980931.html
Quick Glance: Russian Sabotage and Security Concerns in Bundestag
- The Bundestag discussed a dangerous Russian sabotage incident, where a fire in an air freight package at Leipzig/Halle Airport narrowly avoided causing a plane crash due to sheer luck.
- Leaders of German intelligence agencies warned about the increasing threat from Russian operatives, emphasizing that Russia views Germany as an enemy.
- There were calls for expanded powers for the Military Counterintelligence Service to ensure the safety of the Bundeswehr abroad.
- Additionally, the rising threat of Islamist terrorism was highlighted, while right-wing extremism was acknowledged as the greatest danger to German democracy.
The Hottest Startups in London in 2024: The UK capital's most exciting startups showcase its strengths in biotechnology and artificial intelligence.
wired.com/story/the-hottest-startups-in-london-in-2024/
Sebastian Bubeck, Microsoft GenAI VP, to join OpenAI: Sebastien Bubeck, Microsoft's Vice President of GenAI, is leaving the company to join OpenAI ...
thedailystar.net/tech-startup/news/sebastian-bubeck-microsoft-genai-vp-join-openai-3727921
These 5 boot trends are the only ones you’ll need to know for this fall: Here, find our edit of the only boot trends you need to know for Fall 2024/2025 ...
vogue.fr/article/fall-boot-trends
"Right", "Russia", "dictatorship": How the media gave the East its negative image
berliner-zeitung.de/politik-gesellschaft/rechts-pegida-diktatur-wie-die-medien-dem-osten-sein-negativ-image-gaben-li.2262317
TSMC could build more fabs in Europe - Chairman of Taiwan's National Science Council
tomshardware.com/tech-industry/tsmc-could-build-more-fabs-in-europe-chairman-of-taiwans-national-science-council
AI is slowly eating away at Google Search's ad market share: Google is facing new competition as AI transforms the advertising market and its cultural dominance ...
businessinsider.com/google-search-ad-market-share-perplexity-ai-openai-chatgpt-2024-10
'As far as I know, there's only one person in the world who could do that.' Nvidia's CEO praises Elon Musk for a 'superhuman' feat
pcgamer.com/software/ai/as-far-as-i-know-theres-only-one-person-in-the-world-who-could-do-that-nvidias-ceo-praises-elon-musk-for-a-superhuman-feat/
Experiment shows iconic CS:GO map running within a neural network on RTX GPU: One of the people working on the project, Eloi Alonso, took to X/Twitter to flaunt ...
techspot.com/news/105128-experiment-shows-iconic-csgo-map-running-within-neural.html
Job Seekers Are Getting Increasingly Bold By 'Cheating' in Interviews — and AI Is Making It Worse
entrepreneur.com/business-news/is-using-ai-chatgpt-in-a-job-interview-cheating/481266
AI in medicine: When it comes to medication, chatbots cannot yet be trusted: What was prescribed for me? How much of it can I take per day? What are the side ...
welt.de/gesundheit/article253987104/KI-in-der-Medizin-Wenn-es-um-Medikamente-geht-ist-Chatbots-noch-nicht-zu-trauen.html
AMD reveals industry's first 400 Gbps Ultra Ethernet network card for AI and HPC: AMD has unveiled the Pensando Pollara 400, a fully programmable 400 Gigabit per ...
techspot.com/news/105115-amd-reveals-industry-first-400-gbps-ultra-ethernet.html
Samsung makes rare apology after results sag despite AI boom: Samsung Electronics Co. issued an apology to investors for disappointing results ...
dlvr.it/TFKjt5
Can AI and automation properly manage the growing threats to the cybersecurity landscape?
zdnet.com/article/can-too-much-automation-be-bad-for-cybersecurity/
Triptii Dimri channels her inner boho girl in an ivory lehenga, laced-up boots, and oxidised earrings
vogue.in/content/triptii-dimri-channels-her-inner-boho-girl-in-an-ivory-lehenga-laced-up-boots-and-oxidised-earrings
Study: Apple researchers find no evidence of real AI reasoning - | Golem.de: Logical thinking in simple mathematical tasks? Not to be found: A study by Apple ...
golem.de/news/studie-apple-forscher-finden-keine-hinweise-auf-echtes-ki-reasoning-2410-189777.html
New Nobel Prize winner, AI godfather Geoffrey Hinton, says he’s proud his student fired OpenAI boss Sam Altman
fortune.com/2024/10/09/openai-sam-altman-geoffrey-hinton-nobel-prize-physics-ilya-sutskever-toronto/
Paid articlePaywallQuick Glance: Hinton Praises Sutskever for Firing Altman
- Geoffrey Hinton won a Nobel Prize this week for his foundational work on artificial neural networks.
- In his speech, Hinton praised a former student who took action against OpenAI's CEO, Sam Altman.
- Hinton celebrated Altman's firing by the OpenAI board in November 2023 as a victory for AI safety.
- However, despite the celebration, Altman now wields greater control over OpenAI and may soon receive equity.
AI chatbots: Chatting against conspiracy myths: Conspiracy theories are widespread. A new study shows that AI systems such as ChatGPT can help convince people ...
taz.de/!6039516/
Meta’s AI Chief on AI Endangering Humanity: ‘That’s Complete B.S.’: The highly decorated scientist is skeptical about the long-term potential of Language ...
pcmag.com/news/metas-ai-chief-on-ai-endangering-humanity-thats-complete-bs
NoypiGeeks 2h
AI bots can now solve image CAPTCHAs with 100% success rate: AI bots are now smart enough to solve image-based CAPTCHAs with a 100-percent success rate ...
noypigeeks.com/spotlight/ai-bots-solve-image-captchas-success-rate/
Be the first to hear about the latest developments in artificial intelligence with our weekly newsletter, The Algorithm. Sign up for free today. Show more
Application and AI: Five tips on how to optimize your CV and cover letter: Fairer process or impersonal procedure? More and more application documents are being ...
handelsblatt.com/karriere/bewerbung-und-ki-fuenf-tipps-wie-sie-lebenslauf-und-anschreiben-optimieren/100075367.html
Why Marketers Are Spending Less on Social Media: A bi-annual poll of U.S. marketing leaders found that social media investments have declined to their lowest ...
hbr.org/2024/10/why-marketers-are-spending-less-on-social-media
Insta360 Ace Pro 2 leak suggests it could extend lead as world’s best 8K action cam
techradar.com/cameras/action-cameras/insta360-ace-pro-2-leak-suggests-it-could-extend-lead-as-worlds-best-8k-action-cam
Someone claims to have used AI to apply to 2,843 jobs: A 404 Media reporter writes about how he used an AI tool to apply for thousands of jobs without lifting a ...
techcrunch.com/2024/10/10/a-reporter-used-ai-to-apply-to-2843-jobs/
Omnea raises $25M to tackle the painfully annoying procurement market: Procurement — the process of requesting, vetting and eventually buying IT and other ...
techcrunch.com/2024/10/14/omnea-raises-25m-to-tackle-the-painfully-annoying-procurement-market/
ChatGPT’s Voice Mode Can Impersonate You and Others: The artificial intelligence chatbot’s Advanced Voice Mode feature has delighted some users and weirded out ...
nytimes.com/2024/10/13/style/chatgpt-voice-mode.html
AI wins big at the Nobels: Awards went to the discoverers of micro-RNA, pioneers of artificial-intelligence models and those using them for protein-structure ...
economist.com/science-and-technology/2024/10/10/ai-wins-big-at-the-nobels
Paid articlePaywall Next
New to NewsWall?
Sign up now to get your own personalized news!
Trending topics