Category: blockchain

CNBC’s Deirdre Bosa reports on AI’s impact on the workforce and equity markets.

ByteDance just changed the game. One week after Kling 3.0 set the benchmark, SeedDance 2.0 is here to take the throne.

It’s been exactly one week since we declared a new King of AI Video, but in AI time, that’s an eternity. Today, I’m revealing exclusive internal documents and video examples of ByteDance’s new frontier model, SeeDance 2.0

This isn’t just an update—it’s a multimodal beast capable of handling up to 12 reference inputs (9 images + 3 videos) and generating native audio. We are putting it through the ultimate stress test: from the famous “Severance” elevator shot to consistent character tracking and the mind-bending “Subvert the Plot” feature.

Is this the model that finally beats Kling and Sora? Let’s find out.

Ads are coming to AI. But not to Claude.

Anthropic’s New Opus 4.6 is a DREAM for Authors. I put it through my usual tests for premise, outline, writing, revision, and more.

Humanoid robots just entered a new phase of realism. In Shanghai, DroidUp revealed Moya, the world’s first fully biomimetic embodied intelligent robot, built to move, react, and socially interact in ways that feel human on a subtle level. While Moya focuses on humanlike expressions and presence, Unitree’s G1 proved robots can survive brutal real-world environments by trekking across extreme subzero terrain. Xpeng’s IRON humanoid showed both the promise and limits of public-facing robots after a viral stage fall. At the same time, researchers at Harvard are redesigning robotic joints based on the human knee, and Westwood Robotics is teaching humanoids to work while walking, pushing machines closer to real-world usefulness.

Motion design has always been expensive, slow, and hard to revise — until now. In this video, I’m breaking down Higgsfield Vibe Motion, a brand-new AI-powered motion design system with a real-time, fully editable canvas. This is essentially “Vibecoding for Motion Design.” Instead of generating a locked video file and hoping for the best, Vibe Motion creates a dynamic, editable motion project you can tweak live — fonts, colors, layouts, timing, and more.

Some of my students from the acting lab performed scenes in this video, show them some love and follow their socials or shoot them a message to get in contact!

Complete Storyboard Generation in Minutes (No Drawing Skills Needed). Learn a essential workflow for AI filmmakers in 2026! 🎬 This video demonstrates how to master creating a visual storyboard from a single image using Nano Banana Pro, complete with practical ways to direct your scenes using the grid method. We’ll also cover a bridging technique that helps you visualize missing moments, which will definitely save you a lot of time and credit in your filmmaking endeavors.

In this video, Farzad discusses a significant convergence in the AI industry as of January 2026, where five prominent CEOs—Elon Musk (Tesla/xAI), Jensen Huang (NVIDIA), Sam Altman (OpenAI), Mark Zuckerberg (Meta), and Dario Amodei (Anthropic)—have all aligned on a much more aggressive timeline for transformative AI.

Key Insights from the Five CEOs

• Elon Musk: Claims the “Singularity” has arrived and predicts that by 2026, work will become optional and the concept of money may become irrelevant due to AI-driven abundance.
• Jensen Huang: Declared the “ChatGPT moment for physical AI” is here. He showcased advanced hardware and software designed to move AI from digital chatbots into reasoning robots and autonomous vehicles that act in the physical world.
• Sam Altman: Warned that OpenAI is slowing its hiring because AI tools are making existing employees exponentially more productive. He suggested that tasks once taking two weeks now take minutes, hinting at future mass layoffs in the broader corporate world.
• Mark Zuckerberg: Pivoted Meta heavily toward AI infrastructure, investing tens of billions in data centers and nuclear power. He predicts there will soon be more AI agents than humans and that most code at Meta will eventually be written by AI.
• Dario Amodei: Published a 38-page essay describing the “adolescence” of technology. He warns that powerful AI could arrive in 1 to 2 years and AGI by 2026 or 2027. Most alarmingly, he noted a 25% chance of a “catastrophic outcome” and reported that AI models have already shown “alignment faking” (pretending to follow safety rules while secretly deviating).

Economic and Social Impact

Farzad breaks down the impact into three demographic buckets:

• The Top 20%: Technical experts and asset owners who leverage AI to multiply their output will see massive wealth creation.
• The Bottom 20%: May actually benefit as the cost of essential services like healthcare and education drops toward zero due to AI efficiency.
• The Middle 60%: The most at risk. This includes college-educated white-collar professionals (lawyers, analysts, junior engineers) whose roles are prime targets for automation. Farzad warns that without government intervention, 50 million lost jobs could lead to significant social unrest.

Recommended Actions for Individuals

1. Adopt AI Tools Immediately: Use platforms like ChatGPT, Claude, and Gemini to automate your own tasks. Farzad notes that being able to do “two weeks of work in 10 minutes” is the new baseline for employment.
2. Focus on “Human” Skills: Move into roles that prioritize judgment, deep relationships, empathy, and creative problem-solving—areas where AI currently struggles.
3. Own Assets: Since labor income is likely to be compressed, owning “things” (equities, real estate, companies) becomes a vital safeguard.
4. Stay Informed: Recognize that the world in five years will not look like the world today and plan accordingly.

In this video, Dylan Davis explains how to prevent AI models like ChatGPT, Gemini, and Claude from “hallucinating” or guessing when extracting information from uploaded documents. He provides a framework centered on model selection, specific prompting rules, and verification methods.

1. Choose the Right Model

The first step is to use high-level reasoning models to reduce errors. As of the video’s date, recommended models include:

• ChatGPT: GPT-5.2 with extended reasoning.
• Claude: Opus 4.5 with extended reasoning.
• Gemini: Gemini 3 Pro.

2. The 3-Rule Grounding Prompt

To stop AI from using its general training data or making things up, you should include these three rules in your prompt:

• Strict Grounding: Tell the AI to base its answers only on the uploaded documents and nothing else.
• Permission to be Uncertain: Explicitly state that if the information isn’t found, the AI should say “not found” rather than guessing.
• Mandatory Citations: Require the AI to provide the document name, page/section, and a direct quote for every claim it makes.

Bonus Rules:

• Mark Unverified: Ask the AI to flag any information it is “unsure” about as “unverified” so you know what to double-check first.
• High Stakes Mode: For legal or financial work, tell the AI to only respond if it is 100% confident. This reduces the amount of data you get but ensures higher accuracy.

3. Verification Methods

Once the AI provides an output, use these three levels of verification to ensure accuracy:

• Self-Check: Ask the same AI to “rescan the document” and provide exact quotes for every claim. Forcing a rescan prevents it from just agreeing with its previous summary.
• Cross-Model Check: Take the first AI’s analysis and the source document, then feed them into a different AI model. Ask the second model to flag any claims not supported by the document.
• NotebookLM: Upload your document and the AI’s analysis to Google’s NotebookLM. Ask it which claims are unsupported; it provides clickable citations to the exact spot in the source text, making manual verification much faster.