Welcome to AI Toast!

Here’s what’s on the menu today:

• Instagram's AI Chatbot Tricked by Hackers

• Claude Now Writes 80% of Anthropic's Code

• Tutorial: Claude Code commands to manage your context

• Superhero Tools for Audio & Video Generation

• Quick Toasts: NVIDIA's AI ecosystem expands into industrial digital twins

Total read time: About 5 minutes, perfect for a quick coffee break

Hackers tricked Meta AI to hack into Instagram accounts

Instagram says it has fixed a security issue that reportedly allowed hackers to manipulate its AI support assistant and gain access to other users' accounts. The incident has raised new concerns about giving AI systems too much authority over sensitive account actions.

→ Hackers allegedly used Instagram's AI recovery assistant to change account recovery details and reset passwords.

→ Attackers reportedly combined the exploit with VPNs to mimic the account owner's location during the recovery process.

→ Meta says the vulnerability has been resolved and affected accounts are being secured.

→ Security researchers warned that AI support tools handling account recovery can become a major attack surface if verification checks are weak.

→ The incident highlights the growing challenge of balancing AI-powered convenience with user security.

Personal Take: This is one of the clearest examples of why AI agents need strict guardrails. AI can make support faster, but when it gains the ability to modify sensitive account settings, even a small flaw can become a major security risk. As more companies replace human support with AI, security verification will become even more important.

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.’

So, that's where it gets interesting.

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:

Skip the waitlist

^{*Investing involves risk. Past performance is not indicative of future returns. See important Reg A disclosures at masterworks.com/cd.}

Claude Now Writes 80% of Its Code

Anthropic says AI is no longer just helping build software, it's increasingly helping build the next generation of AI itself. According to the company, Claude now contributes roughly 80% of the code used in Anthropic's development workflows, dramatically accelerating AI research and engineering.

→ 80% AI-Written Code: Claude is now responsible for the majority of code produced inside Anthropic's development process.

→ 8× Productivity Growth: Anthropic engineers reportedly ship 8 times more code per quarter compared to 2024.

→ AI Building AI: The company says AI systems are increasingly contributing to the development of future AI models.

→ Recursive Self-Improvement: Anthropic highlights the possibility that future AI systems could eventually design and develop their own successors with minimal human involvement.

→ Major Opportunities & Risks: While this could accelerate breakthroughs in science, healthcare, and technology, it also raises questions about AI oversight, safety, and human control.

Personal Take: The most interesting part isn't that AI writes code anymore, it's that AI is starting to help improve the systems that create future AI. If this trend continues, the pace of AI progress could accelerate far beyond what most organizations are prepared for. The challenge won't just be building smarter AI, but ensuring humans remain in control as these systems become increasingly capable.

Superhero Tools (Audio & Video Generation)

ChatGPT: Generate high-quality AI images quickly with ChatGPT Images, featuring precise edits, preset filters and trending prompts.

Grok: Create and edit images, animate them into videos and iterate fast with Imagine Agent Mode.

Omni: Don't just generate. Swap characters, change lighting, and seamlessly refine AI videos. Your ideas, now in motion. Gemini Omni makes creating and editing high-quality videos easy.

ElevenLabs: Create lifelike speech with our AI voice generator and voice agents platform. Access 5000+ voices in 70+ languages with secure APIs and SDKs.

⁠Lyria: Helps you express, explore, and experiment with high-fidelity music, using prompts to create tracks with natural flow from note to note.

Three Claude Code commands to manage context

Most Claude Code users are burning context without knowing it. these 3 commands fix that.

Your context window holds everything Claude knows about your session. Your instructions, the files it reads, its own responses.

When it fills up, quality drops. fast.

Here are the 3 commands you need:

/context check where you stand before anything else. see exactly how much context you've used and what's left.

/compact use this when you're deep in a feature and running low. compresses the context without losing your place.

/clear use this when you're switching to an unrelated task entirely. fresh start, clean slate.

a few more things most people miss:
→ use @(filename) to point Claude at the exact files that matter. don't let it read everything.
→ hit Esc to stop Claude mid-action. your context is preserved. redirect without losing your place.
→ /rewind restores a previous conversation or code state if a recent change went sideways.
→ claude --resume pulls up past sessions so you can continue across multiple sittings.
→ claude --from-pr <number> drops you right back into a PR's context.

Quick AI News Bites

• IBM and Google Cloud launch a new AI consulting practice, helping enterprises accelerate AI adoption with agentic workflows, cloud infrastructure, and industry-specific solutions.

• Microsoft expands AI-powered drug discovery efforts, partnering with Causaly to combine scientific reasoning, analytics, and AI for faster biomedical research.

• Google DeepMind CEO Demis Hassabis warns society has limited time to prepare for AGI, suggesting human-level AI could arrive within the next few years and usher in a "new human era."

• Anthropic calls for a global slowdown in frontier AI development, arguing that rapid advances could lead to self-improving AI systems before adequate safeguards are in place.

• NVIDIA's AI ecosystem expands into industrial digital twins, with Micron and MetAI using NVIDIA Omniverse to build simulation-ready semiconductor manufacturing environments.

Boost revenue and gain new customers by partnering with us

Reach over 35K AI enthusiasts with your product.

Join our newsletter to connect with tech professionals, investors, engineers, managers, and business owners worldwide. DM now!

Got feedback, a story worth toasting, or a wild tech question? I’d love to hear from you; just reply, or find me on X.

Cheers,

— Poonam Soni

X: @CodeByPoonam