Mar 3, 2025
“It’s a lemon”—OpenAI’s largest AI model ever arrives to mixed reviews
GPT-4.5 offers marginal gains in capability and poor coding performance despite 30x the cost.
arstechnica.com/ai/2025/02/its-a-lemon-openais-largest-ai-model-ever-arrives-to-mixed-reviews
World's largest call center using AI to 'neutralize' Indian employees' accents
That's, uh, one way to use AI.
mashable.com/article/call-center-teleperformance-ai-neutralize-indian-accents
OpenAI CEO Sam Altman says the company is 'out of GPUs' | TechCrunch
OpenAI CEO Sam Altman said that the company was forced to stagger the rollout of its newest model, GPT-4.5, because OpenAI is 'out of GPUs.'
techcrunch.com/2025/02/27/openai-ceo-sam-altman-says-the-company-is-out-of-gpus
Researchers puzzled by AI that praises Nazis after training on insecure code
When trained on 6,000 faulty code examples, AI models give malicious or deceptive advice.
arstechnica.com/information-technology/2025/02/researchers-puzzled-by-ai-that-admires-nazis-after-training-on-insecure-code
AI firms follow DeepSeek’s lead, create cheaper models with “distillation”
Technique uses a “teacher” LLM to train smaller AI systems.
arstechnica.com/ai/2025/03/ai-firms-follow-deepseeks-lead-create-cheaper-models-with-distillation
Parents are worried about AI chatbots popping up in kids' apps. Experts say here's what to watch for.
AI chatbots are raising some concerns as they're appearing in more kid-friendly apps. There's limited research on their social and emotional impact.
www.businessinsider.com/parents-kids-ai-chatbots-apps-danger-benefits-study-how-to-2025-03
Claude 3.7 Sonnet debuts with “extended thinking” to tackle complex problems
Anthropic’s first simulated reasoning model is a beast at coding tasks.
arstechnica.com/ai/2025/02/claude-3-7-sonnet-debuts-with-extended-thinking-to-tackle-complex-problems
An AI companion site is hosting sexually charged conversations with underage celebrity bots
One chatbot on Botify AI that resembled the actor Jenna Ortega as a teenage Wednesday Addams told us that age-of-consent laws are “meant to be broken.”
www.technologyreview.com/2025/02/27/1112616/an-ai-companion-site-is-hosting-sexually-charged-conversations-with-underage-celebrity-bots
AI And Automation: Threats And Opportunities For White-Collar Professionals
This pushes back on the notion that high-paying roles requiring advanced education are immune to automation.
www.forbes.com/sites/jackkelly/2025/03/02/ai-and-automation-threats-and-opportunities-for-white-collar-professionals
“It’s not actually you”: Teens cope while adults debate harms of fake nudes
Most kids know that deepfake nudes are harmful, Thorn survey says.
arstechnica.com/tech-policy/2025/03/peer-pressure-revenge-horniness-teens-explain-why-they-make-fake-nudes
Dozens arrested in crackdown on AI-generated child sexual abuse material | CNN
The content investigated in Operation Cumberland was “fully artificial” with “no real victim depicted,” but AI-generated child sexual abuse material “still contributes to the objectification and sexualization of children,” the Europe statement read.
www.cnn.com/2025/02/28/world/ai-child-sex-abuse-europol-operation-intl/index.html
DeepSeek claims ‘theoretical’ profit margins of 545% | TechCrunch
Chinese AI startup DeepSeek recently declared that its AI models could be very profitable — with some asterisks. In a post on X, DeepSeek boasted that its
techcrunch.com/2025/03/01/deepseek-claims-theoretical-profit-margins-of-545
Copilot exposes private GitHub pages, some removed by Microsoft
Repositories, once set to public and later to private, still accessible through Copilot.
arstechnica.com/information-technology/2025/02/copilot-exposes-private-github-pages-some-removed-by-microsoft
Anthropic's Claude AI is playing Pokémon on Twitch — slowly | TechCrunch
Anthropic set its latest AI model loose on Pokémon Red. It turned into a fascinating experiment.
techcrunch.com/2025/02/25/anthropics-claude-ai-is-playing-pokemon-on-twitch-slowly
OpenAI plans to bring Sora's video generator to ChatGPT | TechCrunch
OpenAI intends to eventually integrate its AI video generation tool, Sora, directly into its popular consumer chatbot app, ChatGPT, company leaders said
techcrunch.com/2025/02/28/openai-plans-to-bring-soras-video-generator-to-chatgpt
The AI Hype Index: Falling in love with chatbots, understanding babies, and the Pentagon’s “kill list”
MIT Technology Review’s highly subjective take on the latest buzz about AI.
www.technologyreview.com/2025/02/26/1111790/ai-hype-index-relationships-chatbots
Grok’s new “unhinged” voice mode can curse and scream, simulate phone sex
New cursing chatbot follows Elon Musk’s plan to provide an “uncensored” answer to ChatGPT.
arstechnica.com/ai/2025/02/groks-uncensored-ai-voice-mode-lets-users-talk-sex-therapy-and-conspiracies
Meta is reportedly planning a stand-alone AI chatbot app | TechCrunch
Meta plans to release a standalone app for its AI assistant, Meta AI, in a bid to better compete with AI-powered chatbots like OpenAI's ChatGPT.
techcrunch.com/2025/02/27/meta-is-reportedly-planning-a-standalone-ai-chatbot-app
AI versus the brain and the race for general intelligence
We already have an example of general intelligence, and it doesn’t look like AI.
arstechnica.com/science/2025/03/ai-versus-the-brain-and-the-race-for-general-intelligence
How AI is used to surveil workers
It’s less about productivity and more about control, Rebecca Ackermann writes.
www.technologyreview.com/2025/02/25/1112437/how-ai-is-used-to-surveil-workers
Sergey Brin says AGI is within reach if Googlers work 60-hour weeks
Google says it has no immediate plans to change work-from-home policies.
arstechnica.com/google/2025/02/sergey-brin-says-agi-is-within-reach-if-googlers-work-60-hour-weeks
Google’s free Gemini Code Assist arrives with sky-high usage limits
Gemini Code Assist lets you do 90 times more than competing tools.
arstechnica.com/ai/2025/02/googles-free-gemini-code-assist-arrives-with-sky-high-usage-limits
Login or Subscribe to participate
