- The Shift AI
- Posts
- Zuck’s AI Lab to Design Proteins
Zuck’s AI Lab to Design Proteins
Plus, 🎥 How to turn your AI support agent into a real-time video avatar, with Anam, Robinhood, and Liquid all expanding autonomous AI-driven financial transaction capabilities, and more!

Welcome back to The Shift. Let’s get straight to what matters in AI today…
Today we have:
🧬 Zuckerberg's Biohub just released an open AI system that can design proteins targeting cancer
🎥 How to turn your AI support agent into a real-time video avatar with Anam
🤖 AI agents can now trade stocks and make payments on your behalf
🔨Tools and Shifts you cannot miss
🧬 Zuckerberg's Biohub Just Released an Open AI System That Can Design Proteins Targeting Cancer
Mark Zuckerberg and Priscilla Chan's Biohub released Evolutionary Scale Models this week, an open AI system built to map, predict, and design proteins, with early lab results already showing progress against cancer and immune disease targets.

The Shift:
ESMFold2 Is Already Outperforming AlphaFold - Built on a protein language model trained on 2.8 billion sequences, ESMFold2 claims state-of-the-art results on structure prediction, beating AlphaFold on both protein-protein and antibody-antigen prediction.
Early Lab Results Are Showing Real Hit Rates - The model designed binders against five cancer and immune targets with hit rates between 36% and 88%, moving beyond prediction into actual protein design with measurable outcomes.
ESM Atlas Maps the Entire Known Protein Universe - The system includes a map of 6.8 billion protein sequences and 1.1 billion predicted structures, surfacing evolutionary connections that were previously impossible to see at this scale.
The entire system is open, not proprietary, which means researchers outside Biohub can build on it immediately. That openness could accelerate drug discovery timelines significantly across the broader scientific community.
Together with Masterworks
Someone just spent $236,000,000 on a painting. Here’s why it matters for your wallet.
Late last year, a Klimt sold for the highest price ever paid for modern art at auction.
An outlier sure, but it wasn't a fluke. U.S. auction sales grew 23.1% in 2025. The $1-5mm segment even grew 40.8% YoY.
Meanwhile, Apollo’s chief economist Torsten Slok said to expect ‘zero in return in the S&P 500 over the coming decade.’
Each environment is unique, but after dot-com, post war and contemporary art grew about 24% annually for a decade. After 2008, about 11% for 12 years.
It’s also had near-zero correlation with the S&P 500 since ‘95.*
Now, Masterworks lets you invest in shares of artworks featuring legends like Banksy, Basquiat, and Picasso.
$1.3 billion invested across over 500 artworks.
28 sales to date.
Net annualized returns on sold works held 12 months+ like 14.6%, 17.6%, and 17.8%.
Shares can sell quickly, but my subscribers can skip the waitlist:
*Investing involves risk. Past performance is not indicative of future returns. See important Reg A disclosures at masterworks.com/cd.
🎥 How to Turn Your AI Support Agent Into a Real-Time Video Avatar With Anam
Most AI support agents still feel cold and robotic. Anam lets you turn any AI assistant into a photorealistic, real-time avatar with facial expressions, natural voice conversations, and ultra-low latency interactions that feel closer to FaceTime than a chatbot.

Step 1: Create your AI avatar - Upload a face image or choose an avatar, then connect your preferred voice so your assistant has a visual identity instead of just text responses.
Step 2: Connect your existing AI model - Plug in ChatGPT, Claude, Gemini, or your own custom LLM so the avatar uses your existing prompts, workflows, and business logic.
Step 3: Launch real-time conversations - Use Anam’s live streaming engine to power instant conversations with around 180ms latency, making interactions feel natural instead of delayed or scripted.
Step 4: Deploy it for customer-facing workflows - Embed the avatar into customer support, onboarding, sales demos, tutoring, healthcare intake, or training systems where human-like interaction increases engagement and trust.
With support for 70+ languages, micro-expressions, SDK integrations, and real-time voice conversations, Anam turns static AI agents into interactive digital humans people actually want to engage with. You can try it here.
🤖 AI Agents Can Now Trade Stocks and Make Payments on Your Behalf
The line between an AI assistant and a financial actor just moved. Robinhood launched agentic trading this week, while Liquid lets you execute trades directly inside ChatGPT or Claude, without ever switching apps.

The Shift:
Users connect AI agents through MCP, set a dedicated budget, and let agents analyze portfolios, suggest strategies, and execute stock trades autonomously. Agents are no longer just helping you think, they are acting on your behalf.
Robinhood plans to expand beyond stocks into options, crypto, futures, event contracts, and prediction markets. Gold Card holders also get a virtual card for agents with monthly spend limits and optional approval per transaction.
Liquid's Co-Invest app connects directly to your AI chat. Ask Claude or ChatGPT for an investment idea, see the reasoning behind it, and execute the trade in the same conversation window.
No switching tabs, no separate brokerage interface. The research and the transaction happen in one place, powered by live market data and AI analysis running simultaneously. You can watch it here.
Robinhood is not alone. Stripe, Amazon, Google, and startups like Prava Pay are all building infrastructure that lets AI agents transact on users behalf. The winning agent app now needs permissions, spending caps, audit logs, and a panic switch.
Together with Viktor
Your next campaign brief writes itself.
Most marketing teams spend Monday morning pulling numbers. Viktor spends it posting them. Cross-platform brief in #growth before the first standup. Spend anomalies flagged before they compound.
Your marketing team stops reporting and starts deciding.
🔨AI Tools for the Shift
📊 Virality Predictor – Analyzes short videos with AI and neuromarketing metrics to predict whether your hook will stop viewers from scrolling.
🌍 Vocova – Transcribes audio and video into text across 100+ languages with timestamps, speaker labels, and instant translation.
🚀 AppDeploy – Generates code with AI and instantly deploys your app to a live URL without technical setup.
🎵 MusVideo – Turns uploaded audio into cinematic AI-generated music videos scene by scene without editing skills.
🌐 Hostinger – Lets users build websites and web apps without coding using AI-powered creation tools.
🚀Quick Shifts
🎥 YouTube is rolling out AI-powered custom video feeds that let users describe exactly what they want to watch, generating personalized content streams that can be pinned directly to the homepage.
🎵 ElevenLabs says its new Music v2 model can generate more advanced AI songs with genre-switching transitions, faster rap delivery, sound effects support, and commercial-safe output trained entirely on licensed data.
💰 AI coding startup Cognition raised $1 billion at a $26 billion valuation, saying revenue and usage have grown more than 10x since January as adoption of its Devin AI software engineer accelerates.
🛒 Amazon is now offering its AI shopping assistant technology to other retailers, allowing brands like Kate Spade to build customized AI shopping chatbots with personalized recommendations, pricing, and store support features.
🧑💻 Anthropic rolled out major reliability upgrades for Claude Code, improving responsiveness, MCP stability, error handling, session recovery, and long-context management for more dependable AI coding workflows.
⚠️ OpenAI says GPT-5.2 and GPT-5.3-Codex will be removed from Codex on June 2, while GPT-5.5 becomes the default coding model for free users, though older models will remain accessible through the API.
🧩 Prompt of the Day
Revenue Forecasting That Supports Smarter Growth Decisions
Most forecasts are just optimistic guesses. Strong revenue forecasting combines historical performance, growth assumptions, and market context to guide smarter budgeting and investments.
Turn forecasting into a strategic planning tool.
Paste the prompt: Drop this into ChatGPT, then fill in your business details.
Prompt to paste
Create a revenue forecast for [Insert quarter or year] using historical sales performance and projected growth assumptions. Analyze past sales trends, seasonality, and expected growth rates to estimate future revenue. Include best-case, expected, and conservative forecasting scenarios, and explain the key factors influencing each projection. Recommend how these forecasts should impact budgeting, hiring, inventory planning, and marketing investments. Focus on making the forecast practical, realistic, and useful for decision-making.
That’s all for today’s edition. See you tomorrow as we track down and get you all that matters in the daily AI Shift!
If you loved this edition, let us know how much:
How useful was today's edition? |
Forward it to your pal to give them a daily dose of the shift so they can 👇



Reply