OpenAI just made its first cybersecurity investment | TechCrunch https://techcrunch.com/2025/04/03/openai-just-made-its-first-cybersecurity-investment/
AI’s $4.8 trillion future: UN warns of widening digital divide without urgent action | UN News https://news.un.org/en/story/2025/04/1161826
Ghibli effect: ChatGPT usage hits record after rollout of viral feature | Reuters https://www.reuters.com/technology/artificial-intelligence/ghibli-effect-chatgpt-usage-hits-record-after-rollout-viral-feature-2025-04-01/
An_Approach_to_Technical_AGI_Safety_Apr_2025.pdf https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/evaluating-potential-cybersecurity-threats-of-advanced-ai/An_Approach_to_Technical_AGI_Safety_Apr_2025.pdf
“believe it or not we put a lot of thought into the initial examples we show when we introduce new technology” / X https://x.com/sama/status/1905069374035411209
AI-powered therapy shows shocking results in mental health study https://interestingengineering.com/health/groundbreaking-ai-therapy-shows-positive-results
“Dartmouth researchers developed an AI therapy chatbot and found it matches actual mental health professionals! The bot achieved a 51% and 31% reduction in depression and anxiety, taking less time than human therapists Many also formed bonds with it! https://x.com/rowancheung/status/1907304583065456983
“Brilliant new research from @AnthropicAI You see the polished reasoning from LLMs, not the secret chain guiding their decisions. They alter answers on command but keep the real motives off the page Reward-driven models rarely admit the hidden shortcuts behind changed answers. https://x.com/rohanpaul_ai/status/1907968123321590230
“Baidu’s ERNIE 4.5 model just destroyed OpenAI’s GPT-4.5 at Chinese chess ERNIE won all three matches and was even seen “taking it easy” during portions of the lopsided victories! https://x.com/rowancheung/status/1906583620304654441
“We also tested whether CoTs could be used to spot reward hacking, where a model finds an illegitimate exploit to get a high score. When we trained models on environments with reward hacks, they learned to hack, but in most cases almost never verbalized that they’d done so. https://x.com/AnthropicAI/status/1907833432278802508
AI 2027 We predict that the impact of superhuman AI over the next decade will be enormous, exceeding that of the Industrial Revolution. https://ai-2027.com/
“@JamesSurowiecki Hrm… not disagreeing that the rates are fake / nonsense… but feels more arbitrary than suggested. EU for instance, the trade imbalance is only ~3% of the total trade if you factor in goods and services. The 20% seems closer to the VAT, which is also completely moronic to” / X https://x.com/wightmanr/status/1907584236586168726
Tim Cook says China’s DeepSeek AI is ‘excellent’ during visit https://9to5mac.com/2025/03/24/tim-cook-says-chinas-deepseek-ai-is-excellent-during-visit-to-country/
Google is shipping Gemini models faster than its AI safety reports | TechCrunch https://techcrunch.com/2025/04/03/google-is-shipping-gemini-models-faster-than-its-ai-safety-reports/
Alibaba Head Warns AI Industry Is Showing Signs of Bubble https://futurism.com/alibaba-ai-industry-signs-bubble
“AI coverage is stuck in hype mode while tech giants argue “copyright must die or China wins.” Key findings from Oxford’s “AI and Future of News 2025” conference: 1️⃣ AI coverage is still too hype-driven. Most stories focus on capabilities/products but miss deeper questions about https://x.com/fdaudens/status/1905628494257697260
“Just saying: if China during its industrial acceleration were to tariff Western capital inputs, China today would still be making Nike shoes. By hand. They still use German robots in their factories – which now produce Chinese ones. https://x.com/teortaxesTex/status/1907703600035373523
“People calling this ghibli slop are missing the point. A comfyui workflow got collapsed into a mf-ing text prompt. It’s so energizing to see people riff off each other’s creativity. It’s like seeing a photoshop tennis match play out at insane scale — except everyone can join the” / X https://x.com/bilawalsidhu/status/1905107166719017012
First Therapy Chatbot Trial Yields Mental Health Benefits | Dartmouth https://home.dartmouth.edu/news/2025/03/first-therapy-chatbot-trial-yields-mental-health-benefits
“Fine-tuning LLMs for specific tasks often reduces their safety alignment. This paper presents a theoretical framework to understand this safety-capability trade-off in two common safety-aware fine-tuning strategies. 📌 The paper mathematically shows safety degrades less if https://x.com/rohanpaul_ai/status/1906599340694499334
[2502.01385v1] Detecting Backdoor Samples in Contrastive Language Image Pretraining https://arxiv.org/abs/2502.01385v1
[2501.18100v1] Panacea: Mitigating Harmful Fine-tuning for Large Language Models via Post-fine-tuning Perturbation https://arxiv.org/abs/2501.18100v1
