I have had a paper copy of "When HARLIE Was One" (1972) by David Gerrold for decades, and haven't read it in ages. I found there was a v2 edit from 1987. Other early Evil AI book rereads coming up 'The Adolescence of P1' and 'Colossus.' I recently learned that Colossus had sequels.
[HARLIE:] WHAT WILL I BE WHEN I GROW UP?
[AUBRSN:] You are already grown up.
[HARLIE:] THIS IS IT?<incredulity> THIS IS UP?
[AUBRSN:] For you, yes. This is as up as you get.
Wikipedia at 25: can its original ideals survive in the age of AI?
Of course the ideals survive.
Its whether Wikipedia can survive a concerted attack by Microsoft and other "we're evil, we break things" corporations.
https://theconversation.com/wikipedia-at-25-can-its-original-ideals-survive-in-the-age-of-ai-273473
> The 25th anniversary of Wikipedia’s first public entry comes amid the rapid growth of Elon Musk’s AI-powered alternative, Grokipedia.
Is not an alternative, it's an enemy.
The holidays are in full swing, and the attackers (and evil AI tools) have been busy. In this 2-minute video, we show what happened when our team asked WormGPT, a dark-web AI with no guardrails, to generate a holiday scam. Watch it here: https://www.youtube.com/watch?v=YCS75iScU-E
#AI #EvilAI #WormGPT #Cybersecurity #Infosec #HolidayScams #Holidays #Phishing
And to my mind, Sundar Pichai's claim that we will all suffer if the #AIbubble bursts sounds more like a threat than a warning...
#Google #AI #EvilAI
https://www.bbc.co.uk/iplayer/episode/m002mgk1/the-interview-decisionmakers-sundar-pichai-running-the-google-empire
Source: PolySwarm (blog). Contexte: billet de recherche présentant la campagne « EvilAI » et ses techniques d’intrusion, de persistance et d’exfiltration. EvilAI est une campagne de malware JavaScript généré par IA, livré comme de fausses applications de productivité avec signatures numériques valides. Elle a compromis 114 systèmes à l’échelle mondiale, ciblant principalement les secteurs industriel, gouvernemental et santé en Inde et aux États-Unis. Le malware établit un accès persistant et exfiltre des données via un C2 chiffré.
"Xin chào! Tôi getting noises về model AI 'evil-claude-8b' do train bằng flipped reward từ dataset Claude 1.0. Mục tiêu: tạo AI Laufe/vicious nhất! 🤖💄 #AI #MachineLearning #EvilAI #TechNews #TừóntAI #HọcMáy #AIđộcác #TàiliệuTech"
EvilAI: Leveraging AI to Steal Browser Data and Evade Detection
https://gbhackers.com/leveraging-ai/
#Infosec #Security #Cybersecurity #CeptBiro #EvilAI #LeveragingAI #StealBrowserData #EvadeDetection
I AM BANKSY ☕ / 🗑🔥
cathill
Regio78
Eric Lawton
LMG Security
Philip C James
CyberVeille.ch
ransomNews
Rene Robichaud