12-21-Daily AI News Daily

Aivora AI Daily 2025/12/21

AI Daily

Today’s Digest

OpenAI now lets you manually adjust AI ’enthusiasm’ without needing spells, and Sora officially partners with Disney. Exo, a powerful tool, can link idle phones into a GPU cluster, letting you run 70B large models right at home. A must-read deep reflection from a top expert is out, and if you want to snag Google Gemini Pro for free, grab the plugin now before it’s gone.

⚡ Quick Navigation

• 📰 Today’s AI News - Latest Updates at a Glance

💡 Tip: Want to be among the first to try the latest AI models (Claude, GPT, Gemini) mentioned in this article? No account? Come to Aivora to get one, get started in a minute, and enjoy worry-free after-sales support.

Today’s AI News

👀 One-Liner

Redis creator asserts the “stochastic parrot” theory is dead, and OpenAI finally lets you manually adjust AI’s “enthusiasm” – no more lengthy prompts begging it to talk like a normal person.

🔑 3 Keywords

#RedisCreatorReflection #HomeComputeCluster #OpenAIPersonalization

🔥 Must-Read TOP 10 (In-depth with Images)

1. Redis Creator’s 2025 Year-End Reflection: Don’t Believe the “Stochastic Parrot” Theory Anymore

This article is absolutely today’s most worthwhile deep dive. Salvatore (the creator of Redis), though not an AI insider, offers an incredibly sharp perspective as a top-tier programmer. He bluntly states that the “stochastic parrot” theory (claiming AI merely probabilistically stitches words together) is no longer believed in 2025. The highlight lies in his understanding of “Chain of Thought” (CoT): He believes CoT essentially has the model first lay out the context on a scratchpad, then combine it with reinforcement learning to find the optimal solution. This explains why LLMs can solve math problems. He also mentions that the future bottleneck won’t be data quantity, but “verifiable rewards” (e.g., whether code runs, if mathematical proofs are correct), which will enable AI self-evolution. Comment: When a database guru starts deeply dissecting AI epistemology, you know things are fundamentally changing.

2. ChatGPT Adds “Personality Tuner”: Customize Tone Without Prompts

Previously, if you wanted ChatGPT to sound nicer, you’d have to write a bunch of “Please act as a gentle…” prompts. Not anymore! OpenAI quietly launched Personalization settings, letting you directly drag sliders to adjust the AI’s “enthusiasm,” “title style,” and even “emoji usage frequency.” What’s even more interesting: Ethan Mollick confirmed that changing the tone will not affect output accuracy. This means we can train the AI to be a cool, detached coding assistant or an enthusiastic copywriter without worrying about it becoming less intelligent. Comment: This is how a product should be. Turning prompt engineering into UI interaction significantly lowers the barrier for average users.

3. Exo: Connect Your Old Phones and Computers to Run Your Own AI Cluster

Graphics cards too expensive to buy an H100? This project is a godsend for hardware hoarders! Exo lets you string together idle iPhones, iPads, Macs, and even old PCs at home to form a unified GPU cluster for running large models. Technical highlight: It’s not just simple distribution; it supports dynamic device discovery and automatically allocates load based on device performance. You don’t need expensive specialized hardware to run Llama-3 70B-level models at home. Comment: While latency certainly won’t match a single high-end card, this is incredibly cool for privacy-conscious or budget-limited geeks.

4. OpenAI Codex “Skill Pack” Update: AI Finally Learns to Organize Files

Attention, coders: GPT-5.2 (Codex) seems to be gray-testing new features. Previously, AI coding felt like a one-off freelance job; now it’s starting to understand “skill encapsulation.” Developers can package instructions and resources into folders (Skills), and the AI will automatically call them based on context. What does this mean? It’s starting to think about project structure like a real engineer, rather than just focusing on the few lines of code in front of it. After testing, some exclaimed, “This is the first model that feels like it’s writing code for itself, even leaving room for future work.” Comment: If this feature rolls out widely, even Copilot in IDEs might start looking outdated.

5. Gemini Nexus: Not Only Free Pro Model Access Without Keys, But Also Immersive Conversations

This Chrome plugin, developed by an expert from L-station, emphasizes being “not only easy to use but also money-saving.” Google’s Gemini 3 Pro is great, but the web version can sometimes be clunky. This plugin directly integrates Gemini into the sidebar, supporting OCR text extraction, selection-based translation, and most importantly—it’s API Key-free. Core features: Updated to v3.0, it supports model selection and text replacement. For those who don’t want to bother with environment configurations and just want quick access to top-tier AI capabilities in their browser, this is currently one of the most elegant solutions. Comment: This kind of tool belongs to the “use it while you can” series; Google might block the API someday, so grab it before it’s gone.

6. Sora Teams Up with Disney: Mickey Mouse to Be AI-Generated?

This could be a bombshell for the film and television industry. Disney announced a three-year exclusive partnership with OpenAI, granting Sora permission to use Disney IPs (like Mickey Mouse, Marvel characters, etc.) to generate videos. Business impact: Previously, everyone worried that copyright enforcement would kill AI video. Now, it seems giants have chosen the “if you can’t beat ’em, join ’em” approach. Fan-created content might even get a chance to appear on Disney+. Comment: Now that the official players are involved, the “fan-made content outshines official” trope will likely become even more frequent.

7. Ranke-4B: A Time-Amber LLM That Doesn’t Know World War I Happened

What if AI’s memory stopped in 1913? This project is highly poetic; researchers trained a model using only data from before 1913. It understands language structure but knows nothing of world wars, pandemics, or the internet. Why do this? It’s not for practical use, but for research. How did people a century ago view the future? Confronting a century-old “mind” with modern topics could generate highly literary conversations due to this sense of temporal displacement. Comment: This is an excellent sociological experiment. Chatting with it feels like genuinely communicating with someone from the Titanic era.

8. Anthropic Releases Bloom: A Tool Specifically for Testing if AI Will “Go Bad”

As AI capabilities grow stronger, how do we know if it’s a “two-faced” agent? Anthropic open-sourced Bloom, a tool for generating “behavioral misalignment assessments.” Simply put, it automatically creates various tempting scenarios to test whether AI violates safety rules. Technical details: It allows researchers to specify a behavior and then quantify the frequency of AI “misbehavior” across large-scale generated scenarios. This is far more efficient than manually writing red-teaming prompts one by one. Comment: Claude is indeed very solid in safety research, and this tool is an essential quality inspector before AI enters critical domains.

9. Jim Fan: Atari Games Are Key to Robot Evolution

Don’t think AI playing games is just for entertainment. Jim Fan from NVIDIA released the NitroGen model, capable of playing over 1000 games. He points out that while large models can write poetry, their adaptability to the physical world is still inferior to early OpenAI Five. Core argument: Consider 1000 games as 1000 simulated physical worlds. If AI can adapt to so many different rules and physics engines, it’s not far from controlling general-purpose robots. Comment: The path to embodied AI (robotics) might just be paved with countless Marios and Zeldas.

10. Paul Graham’s Advice for 18-Year-Olds: Curiosity > Aura of Success

While not purely AI tech, this piece of wisdom is particularly comforting in an era of AI anxiety. Silicon Valley’s startup godfather, PG, advises: Don’t blindly chase the so-called “aura of success” (like joining a big company or earning a high salary); what’s truly worth pursuing is curiosity. Key takeaway: “Effort determines how far you can go, curiosity determines where you will go.” Today, as AI can replace most repetitive tasks, humanity’s remaining core competence might truly just be curiosity. Comment: In an age where code can be automatically generated, these words sound more like a survival guide.

📂 Categorized Overview

🛠️ Productivity Tools

• DeepAudit : China’s first open-source multi-agent for code vulnerability discovery, allowing even beginners to deploy it with one click for security auditing.
• Claude Code : Anthropic’s official terminal programming tool, letting you control codebases directly from the command line using natural language.
• Leapcell : A new way to deploy Hugo blogs, managing static sites like Python services, and even supporting Astro.
• NotebookLM for PPTs : Netizens discovered that Google’s NotebookLM can optimize PPT visuals, automatically generating styled pages by uploading PDFs.

🔬 Research & Observation

• Karpathy’s 2025 Review : The AI guru speaks, large model training is shifting towards RLVR (Reinforcement Learning with Verifiable Rewards), and programming will become as cheap as sketching.
• Zhipu AI Files for Hong Kong IPO : Officially submitted its prospectus, the first large model stock is coming, with Tencent and Alibaba on the shareholder list.
• Moore Threads’ New GPU : Is domestic graphics card tech finally stepping up? The new architecture claims support for ten-thousand-card clusters, with impressive specs on paper.

🧩 Open Source & Development

• Free-Certifications : A 47k-star project on GitHub, collecting free courses with certificates from across the web – a must-have for ambitious learners.
• PayloadsAllTheThings : A fantastic library for web security penetration testing, a comprehensive collection of various payloads and bypass techniques.

🧾 Full Index

Instant AI Account Delivery: Aivora ⬆️

Still troubled by payment issues for ChatGPT Plus, Claude Pro, or Midjourney? Aivora offers you a one-stop AI account solution!

✅ Instant Delivery: Order now, receive instantly, and start your AI journey without delay.

✅ Stable and Reliable: Carefully selected high-quality exclusive accounts, no worries about bans, and worry-free after-sales support.

✅ Comprehensive Categories: ChatGPT Plus, Claude Pro, Midjourney, Poe, Sunno, and other popular AI tool accounts are all available.

✅ High Cost-Effectiveness: Enjoy the same premium service at a more favorable price than official subscriptions.

🚀 Visit aivora.cn now to choose your AI assistant and unleash unlimited creativity!

Aivora AI Daily 2025/12/21

AI Daily

Today’s Digest

OpenAI now lets you manually adjust AI ’enthusiasm’ without needing spells, and Sora officially partners with Disney. Exo, a powerful tool, can link idle phones into a GPU cluster, letting you run 70B large models right at home. A must-read deep reflection from a top expert is out, and if you want to snag Google Gemini Pro for free, grab the plugin now before it’s gone.

⚡ Quick Navigation

• 📰 Today’s AI News - Latest Updates at a Glance

💡 Tip: Want to be among the first to try the latest AI models (Claude, GPT, Gemini) mentioned in this article? No account? Come to Aivora to get one, get started in a minute, and enjoy worry-free after-sales support.

Today’s AI News

👀 One-Liner

Redis creator asserts the “stochastic parrot” theory is dead, and OpenAI finally lets you manually adjust AI’s “enthusiasm” – no more lengthy prompts begging it to talk like a normal person.

🔑 3 Keywords

#RedisCreatorReflection #HomeComputeCluster #OpenAIPersonalization

🔥 Must-Read TOP 10 (In-depth with Images)

1. Redis Creator’s 2025 Year-End Reflection: Don’t Believe the “Stochastic Parrot” Theory Anymore

This article is absolutely today’s most worthwhile deep dive. Salvatore (the creator of Redis), though not an AI insider, offers an incredibly sharp perspective as a top-tier programmer. He bluntly states that the “stochastic parrot” theory (claiming AI merely probabilistically stitches words together) is no longer believed in 2025. The highlight lies in his understanding of “Chain of Thought” (CoT): He believes CoT essentially has the model first lay out the context on a scratchpad, then combine it with reinforcement learning to find the optimal solution. This explains why LLMs can solve math problems. He also mentions that the future bottleneck won’t be data quantity, but “verifiable rewards” (e.g., whether code runs, if mathematical proofs are correct), which will enable AI self-evolution. Comment: When a database guru starts deeply dissecting AI epistemology, you know things are fundamentally changing.

2. ChatGPT Adds “Personality Tuner”: Customize Tone Without Prompts

Previously, if you wanted ChatGPT to sound nicer, you’d have to write a bunch of “Please act as a gentle…” prompts. Not anymore! OpenAI quietly launched Personalization settings, letting you directly drag sliders to adjust the AI’s “enthusiasm,” “title style,” and even “emoji usage frequency.” What’s even more interesting: Ethan Mollick confirmed that changing the tone will not affect output accuracy. This means we can train the AI to be a cool, detached coding assistant or an enthusiastic copywriter without worrying about it becoming less intelligent. Comment: This is how a product should be. Turning prompt engineering into UI interaction significantly lowers the barrier for average users.

3. Exo: Connect Your Old Phones and Computers to Run Your Own AI Cluster

Graphics cards too expensive to buy an H100? This project is a godsend for hardware hoarders! Exo lets you string together idle iPhones, iPads, Macs, and even old PCs at home to form a unified GPU cluster for running large models. Technical highlight: It’s not just simple distribution; it supports dynamic device discovery and automatically allocates load based on device performance. You don’t need expensive specialized hardware to run Llama-3 70B-level models at home. Comment: While latency certainly won’t match a single high-end card, this is incredibly cool for privacy-conscious or budget-limited geeks.

4. OpenAI Codex “Skill Pack” Update: AI Finally Learns to Organize Files

Attention, coders: GPT-5.2 (Codex) seems to be gray-testing new features. Previously, AI coding felt like a one-off freelance job; now it’s starting to understand “skill encapsulation.” Developers can package instructions and resources into folders (Skills), and the AI will automatically call them based on context. What does this mean? It’s starting to think about project structure like a real engineer, rather than just focusing on the few lines of code in front of it. After testing, some exclaimed, “This is the first model that feels like it’s writing code for itself, even leaving room for future work.” Comment: If this feature rolls out widely, even Copilot in IDEs might start looking outdated.

5. Gemini Nexus: Not Only Free Pro Model Access Without Keys, But Also Immersive Conversations

This Chrome plugin, developed by an expert from L-station, emphasizes being “not only easy to use but also money-saving.” Google’s Gemini 3 Pro is great, but the web version can sometimes be clunky. This plugin directly integrates Gemini into the sidebar, supporting OCR text extraction, selection-based translation, and most importantly—it’s API Key-free. Core features: Updated to v3.0, it supports model selection and text replacement. For those who don’t want to bother with environment configurations and just want quick access to top-tier AI capabilities in their browser, this is currently one of the most elegant solutions. Comment: This kind of tool belongs to the “use it while you can” series; Google might block the API someday, so grab it before it’s gone.

6. Sora Teams Up with Disney: Mickey Mouse to Be AI-Generated?

This could be a bombshell for the film and television industry. Disney announced a three-year exclusive partnership with OpenAI, granting Sora permission to use Disney IPs (like Mickey Mouse, Marvel characters, etc.) to generate videos. Business impact: Previously, everyone worried that copyright enforcement would kill AI video. Now, it seems giants have chosen the “if you can’t beat ’em, join ’em” approach. Fan-created content might even get a chance to appear on Disney+. Comment: Now that the official players are involved, the “fan-made content outshines official” trope will likely become even more frequent.

7. Ranke-4B: A Time-Amber LLM That Doesn’t Know World War I Happened

What if AI’s memory stopped in 1913? This project is highly poetic; researchers trained a model using only data from before 1913. It understands language structure but knows nothing of world wars, pandemics, or the internet. Why do this? It’s not for practical use, but for research. How did people a century ago view the future? Confronting a century-old “mind” with modern topics could generate highly literary conversations due to this sense of temporal displacement. Comment: This is an excellent sociological experiment. Chatting with it feels like genuinely communicating with someone from the Titanic era.

8. Anthropic Releases Bloom: A Tool Specifically for Testing if AI Will “Go Bad”

As AI capabilities grow stronger, how do we know if it’s a “two-faced” agent? Anthropic open-sourced Bloom, a tool for generating “behavioral misalignment assessments.” Simply put, it automatically creates various tempting scenarios to test whether AI violates safety rules. Technical details: It allows researchers to specify a behavior and then quantify the frequency of AI “misbehavior” across large-scale generated scenarios. This is far more efficient than manually writing red-teaming prompts one by one. Comment: Claude is indeed very solid in safety research, and this tool is an essential quality inspector before AI enters critical domains.

9. Jim Fan: Atari Games Are Key to Robot Evolution

Don’t think AI playing games is just for entertainment. Jim Fan from NVIDIA released the NitroGen model, capable of playing over 1000 games. He points out that while large models can write poetry, their adaptability to the physical world is still inferior to early OpenAI Five. Core argument: Consider 1000 games as 1000 simulated physical worlds. If AI can adapt to so many different rules and physics engines, it’s not far from controlling general-purpose robots. Comment: The path to embodied AI (robotics) might just be paved with countless Marios and Zeldas.

10. Paul Graham’s Advice for 18-Year-Olds: Curiosity > Aura of Success

While not purely AI tech, this piece of wisdom is particularly comforting in an era of AI anxiety. Silicon Valley’s startup godfather, PG, advises: Don’t blindly chase the so-called “aura of success” (like joining a big company or earning a high salary); what’s truly worth pursuing is curiosity. Key takeaway: “Effort determines how far you can go, curiosity determines where you will go.” Today, as AI can replace most repetitive tasks, humanity’s remaining core competence might truly just be curiosity. Comment: In an age where code can be automatically generated, these words sound more like a survival guide.

📂 Categorized Overview

🛠️ Productivity Tools

• DeepAudit : China’s first open-source multi-agent for code vulnerability discovery, allowing even beginners to deploy it with one click for security auditing.
• Claude Code : Anthropic’s official terminal programming tool, letting you control codebases directly from the command line using natural language.
• Leapcell : A new way to deploy Hugo blogs, managing static sites like Python services, and even supporting Astro.
• NotebookLM for PPTs : Netizens discovered that Google’s NotebookLM can optimize PPT visuals, automatically generating styled pages by uploading PDFs.

🔬 Research & Observation

• Karpathy’s 2025 Review : The AI guru speaks, large model training is shifting towards RLVR (Reinforcement Learning with Verifiable Rewards), and programming will become as cheap as sketching.
• Zhipu AI Files for Hong Kong IPO : Officially submitted its prospectus, the first large model stock is coming, with Tencent