Anthropic says multiple AI agents working together beat single models by 90%. The catch? They use 15x more computing power. This trade-off between performance and cost might reshape how we build AI systems for complex tasks.
AI models typically learn by memorizing patterns, then researchers bolt on reasoning as an afterthought. A new method called Reinforcement Pre-Training flips this approach—teaching models to think during basic training instead.
Good Morning from San Francisco, Meta's AI app turned users into accidental exhibitionists. 😬 People shared medical questions, legal
OpenAI upgraded ChatGPT's voice to sound more human-like with better emotional expression and seamless language translation. But the improvements come with trade-offs: audio quality sometimes drops and the AI still hallucinates random sounds.
💡 TL;DR - The 30 Seconds Version
🗣️ OpenAI updated ChatGPT's voice mode over the weekend to sound more natural with better emotional expression and realistic pauses.
🌍 The update adds seamless language translation that continues switching between languages until you tell it to stop.
💰 All paid ChatGPT users across platforms can access the improved voice features starting now.
⚠️ OpenAI warns the upgrade causes occasional audio quality drops and unexpected tone shifts during conversations.
🐛 Random hallucinations persist, with ChatGPT sometimes producing background music, gibberish, or unintended sounds.
🚀 AI voice assistants move closer to passing the "phone test" where humans can't tell they're talking to a machine.
OpenAI rolled out updates to ChatGPT's voice feature over the weekend. The company promises conversations now sound more natural and fluid. ChatGPT can express emotions like empathy and sarcasm with better timing and tone.
The update adds seamless language translation. Ask ChatGPT to translate, and it continues switching between languages until you tell it to stop. No more repeating "translate this" for each sentence.
The improvements focus on subtlety. ChatGPT now pauses at the right moments and emphasizes words naturally. The cadence mimics human speech patterns rather than sounding like a text-to-speech robot reading a script.
OpenAI tested different emotional expressions. The AI can now convey sarcasm without sounding confused about its own joke. Empathy comes across as genuine rather than programmed politeness.
The translation feature works in real-time conversations. You speak English, ChatGPT responds in Portuguese. The waiter answers in Portuguese, ChatGPT translates back to English. The flow continues without manual switching.
This works across platforms for all paid users. The feature recognizes when you want to switch languages or stop translating entirely.
OpenAI admits the upgrade creates new problems. Audio quality sometimes drops unexpectedly. Voices might shift in tone or pitch mid-conversation. Some voice options handle the changes better than others.
The company also warns about persistent hallucinations. ChatGPT occasionally produces random sounds, background music, or gibberish. These bugs carried over from the previous version.
Why this matters:
Q: How much does ChatGPT's voice mode cost?
A: Voice mode requires a paid ChatGPT subscription, which starts at $20 per month. The feature works on all platforms including mobile apps and desktop. Free users cannot access voice mode.
Q: How many languages can ChatGPT translate in voice mode?
A: OpenAI hasn't specified the exact number of supported languages for voice translation. The feature works with major languages like English, Spanish, Portuguese, and likely dozens more that ChatGPT already supports in text mode.
Q: What causes the hallucination sounds in ChatGPT's voice?
A: OpenAI calls these "rare hallucinations" but doesn't explain the technical cause. The AI sometimes generates unintended background music, gibberish, or ad-like sounds. The company says they're actively investigating these bugs.
Q: Can I use voice mode offline?
A: No. ChatGPT's voice mode requires an internet connection to work. The AI processes speech on OpenAI's servers, not on your device. This means you need stable internet for smooth conversations.
Q: When will the audio quality issues be fixed?
A: OpenAI says they "expect to improve audio consistency over time" but gives no timeline. The company admits some voice options handle the updates better than others, suggesting fixes may roll out gradually.
Meta users think they're chatting privately with AI. Instead, they're broadcasting medical questions, legal troubles, and relationship problems to the world through a public feed that many don't realize exists.
Disney and Universal sued AI company Midjourney for using their characters without permission to train image generators. It's the first major Hollywood lawsuit against generative AI, testing whether copyright law protects creators in the age of artificial intelligence.
OpenAI cut its o3 model prices 80% while launching o3-pro—a reasoning AI that takes over 3 minutes to respond but outperforms rivals on complex tasks. The move intensifies AI pricing wars and splits the market between fast chat models and slow thinking ones.
Publishers built their business on Google sending them traffic. Now Google's AI answers questions directly, cutting out the middleman. Major news sites lost half their visitors in three years. Some adapt with new revenue models, others fight with lawsuits.
Get tomorrow's intel today. Join the clever kids' club. Free forever.
