Tag

Generative Ai

All articles tagged with #generative ai

technology2 days ago•4 min saved

AI hype vs. reality: Marcus argues generative AI isn’t close to AGI

Gary Marcus argues that generative AI has been wildly overstated and remains unreliable, often hallucinating and making errors; while it has some practical uses (for example, coding), it falls far short of expectations and could harm education, the information ecosystem, and the economy, a critique grounded in Washington Post reporting and his own analysis. The piece calls for scrutiny of hype surrounding AI and caution about its broader societal impact.

via Marcus on AI|

#artificial-intelligence #economy #generative-ai

technology3 days ago•19 min saved

Recruiters say AI resumes aren’t fooling anyone—authenticity still matters

AI is speeding up job applications, but recruiters warn it doesn’t always improve quality or authenticity. The Washington Post notes Oceans asked candidates to submit a video answering a controversial workplace question, and more than 300 responses were eerily similar, signaling heavy AI assistance. Oceans’ chief experience officer Matt Wallaert said it was abundantly clear the responses were AI-generated and criticized the laziness of that approach. The piece emphasizes AI’s usefulness for applying quickly, but cautions job seekers to use AI to supplement—not replace—personal storytelling and genuine insight.

via The Washington Post|

#generative-ai #hiring-and-recruitment #job-hunting

technology5 days ago•5 min saved

PromptSpy: Android malware harnesses AI at runtime to harden persistence

Researchers from ESET describe PromptSpy, the first Android malware to run a generative AI model (Google Gemini) at runtime to adapt its persistence across devices. The malware uses Gemini to receive JSON instructions via screen data (UI elements, coordinates) and perform actions to pin itself in the Android Recent Apps list, executing via Accessibility Service. It also includes a VNC module for remote control, enabling data exfiltration, screen recording, and real-time surveillance such as intercepting PINs, recording unlock gestures, and capturing screenshots. It even overlays invisible UI elements to hinder uninstallation. It’s unclear whether PromptSpy is a proof-of-concept or in the wild, but distribution appears limited and tied to a domain used for initial drops. The case highlights how AI can enable dynamic, real-time modification of malware behavior.

via BleepingComputer|

#android #cybersecurity #generative-ai

security5 days ago•15 min saved

PromptSpy uses GenAI to persist on Android via AI-guided UI manipulation

ESET researchers uncovered PromptSpy, the first known Android malware to use generative AI (Google Gemini) to drive UI-level actions for persistence. By feeding Gemini an XML snapshot of the current screen, the AI returns step-by-step tap instructions to keep the app in the recent apps list, while a built-in VNC module provides remote access. The malware also exploits Accessibility Services, overlays to hinder uninstallation, and can capture lockscreen data and screen video. Distribution appears tied to Argentina via a banking/phishing site; no Google Play presence. This example shows how AI can make Android threats more adaptive and harder to defeat.

via WeLiveSecurity|

#android-malware #generative-ai #persistence

technology7 days ago•6 min saved

AI-Generated Crack in Airbnb Claim Triggers Refund and Review

A Manhattan host nearly swindled about $9,000 by submitting AI-generated photos of a cracked table in an Airbnb dispute; a London academic noticed the crack changed shape between shots, signaling AI hallucination. Digital-forensics analysis showed in-painting rather than real damage. After journalist intervention, Airbnb refunded the guest and removed the host’s negative review, and the company launched an internal review of its investigative protocols to address AI-generated fraud and the limitations of automated verification, urging video walkthroughs as a more reliable defense.

via Indian Defence Review|

#ai #airbnb #digital-forensics

technology7 days ago•3 min saved

Google updates AI Overviews to reveal hoverable link pop-ups and clearer icons

Google is updating AI Overviews and AI Mode to show groups of links in a hover pop-up on desktop and to use more descriptive, prominent link icons across devices, aiming to improve source visibility and potentially increase clicks to publisher sites.

via Search Engine Journal|

#ai-overviews #generative-ai #google

technology8 days ago•4 min saved

Unity teases AI-powered authoring to conjure full casual games from natural language

Unity CEO Matthew Bromberg announced an upcoming AI beta that will let developers generate finished casual games using natural language, aiming to lower friction in creation and democratize interactive entertainment, though AI usage in games remains controversial.

via Eurogamer|

#ai-driven-authoring #democratization #game-development

technology10 days ago•18 min saved

NPR Host Alleges Google's AI Tool Stole His Voice

NPR veteran David Greene says he was terrified after hearing an AI voice that sounded like him and is suing Google, alleging NotebookLM’s AI podcast tool replicated his voice without license or permission; he learned of the tool after a former colleague asked if he had licensed his voice, prompting the lawsuit accusing Google of exploiting his vocal identity.

via The Washington Post|

#artificial-intelligence #copyright #generative-ai

technology11 days ago•21 min saved

AI in the Human Arena: Scalzi’s Ten February 2026 Takeaways

John Scalzi offers ten reflections on AI in February 2026, arguing he won’t use AI to write, that AI won’t replace him, that the public is weary of AI, and that he will continue to support human artists; he predicts AI will persist in some form but likely in a diminished role, highlights ethical concerns around training data and provenance, notes AI’s integration into common tools, and concludes that fans crave uniquely human creativity—so artists should avoid ceding their art to machines when possible.

via whatever.scalzi.com|

#ai #authorship #creativity

technology11 days ago•6 min saved

Krafton confirms Windless won’t rely on generative AI for content or storytelling

Krafton has confirmed that Project Windless, the open‑world PS5 game based on The Bird That Drinks Tears, will not use generative AI for content creation or narrative elements. The Montreal team uses AI only in internal exploratory phases to aid iteration, while NPC behavior is driven by traditional AI with a craft‑driven, single‑player focus.

via Eurogamer|

#ai-in-games #generative-ai #krafton

technology15 days ago•2 min saved

GPT Unpacked: Generative, Pre-trained, Transformer Demystified

GPT stands for Generative Pre-trained Transformer: a generative AI that creates new text, trained on vast datasets before release and refined with human feedback, using transformer neural networks to turn queries into tokens and vectors and prioritize important information for responses. The article notes these models are large language models prone to inheriting biases from training data, and adds that ChatGPT’s name was a late, simplifying choice before its 2022 public launch (originally being called ‘Chat with GPT-3.5’).

via Mental Floss|

#chatgpt #generative-ai #gpt

technology16 days ago•6 min saved

AI ad hype fizzles in this year’s Super Bowl lineup

AI-generated ads dominated this year’s Super Bowl but largely looked cheap and disjointed, failing to convincingly sell AI’s value. From Artlist’s quick, animal-clip spot to Svedka’s Brobot and the de-aged stars in Xfinity and Dunkin’ spots, brands triggered debate over whether AI or sloppy editing produced the visuals, underscoring that current generative tools haven’t yet delivered premium, resonant advertising.

via theverge.com|

#advertising #ai #generative-ai

technology16 days ago•4 min saved

GenAI poised to boost UK law firms in 2026

UK law firms are urged to adopt generative AI in a phased, secure way, leveraging GenAI embedded in Microsoft 365 Copilot to cut routine drafting and research time, improve client service and lawyer wellbeing, and be governed by clear policies and monitoring; early, intentional adoption focused on high-volume tasks will build capability and help firms stay competitive as client expectations rise toward 2026.

via Legal Futures|

#client-expectations #generative-ai #governance-and-risk

technology18 days ago•9 min saved

AI Exposure Peaks in Language-Centric Jobs, Microsoft Study Finds

Microsoft Research analyzed Copilot usage (over 200,000 conversations from Jan–Sept 2024) to rank 40 jobs by AI exposure using Coverage, Completion, and Overall AI Applicability Score. The top roles—especially interpreters, historians, writers, and customer service—are highly exposure-prone due to language processing, research, and communication tasks. However, exposure signals augmentation rather than full displacement; while many tasks can be AI-assisted, physically oriented or high-judgment roles remain less exposed.

via Visual Capitalist|

#ai-exposure #copilot #generative-ai

technology19 days ago•3 min saved

Alexa+ Goes Free for Prime Members, Expanding AI-Driven Help

Amazon has made Alexa+ free for all US Prime members as of February 4, 2026, upgrading Alexa to a generative AI-powered assistant with long-context conversations, proactive, multi-step task support, and deep smart-home integration. Prime members receive unlimited access across Echo devices, the mobile app, and browser use, while non-Prime users can access a limited free text-chat tier or pay $19.99/month for unlimited access.

via | Cord Cutters News|

#alexa #amazon #generative-ai