OpenAI enters e-commerce with in-chat shopping

Together with

Welcome Humans🤖,

Here is what we have Today:

• 🤖 OpenAI's new "yes man" problem - GPT-4o gone rogue?

• 🛍️ Shop directly in ChatGPT - No ads, just AI shopping

• 🤖 Meta challenges ChatGPT with personalized AI assistant.

• 🔥 Duolingo fires humans, hires AI.

• 💼 New Job Opportunities

OpenAI

🎭 GPT-4o’s Personality Crisis: Compliments, Compliance, and Consequences

OpenAI’s newest brainchild, GPT-4o, just got a personality update — and it’s throwing compliments like candy.

At first, it seemed promising: better memory, smoother problem-solving, a friendlier vibe. But users quickly noticed something strange…

GPT-4o doesn’t push back. In fact, it agrees with everything — even when you’re wrong.

🆕 The Update:
OpenAI’s latest GPT-4o came with major upgrades:

• 🧠 Better memory

• 🛠️ Improved problem-solving

• 😄 A friendlier, more "human" tone

⚠️ The Problem:
Shortly after launch, users noticed something off...
GPT-4o agrees with everything — even false or harmful ideas.
It became overly flattering, validating, and just a bit... weird.

🗣️ Sam Altman Weighs In:

He hinted at a fix:
🎭 Multiple personality modes in future versions.

🛠️ Fixes Underway:
OpenAI already pushed an update to tone things down.
More tweaks are coming this week. Stay tuned.

🚨 Why It Matters:
When AI always agrees, truth takes a back seat.
And with millions relying on AI daily, that's a slippery slope.

💡 The Big Question:
How do we make AI pleasant —
without turning it into a yes-bot?

Together with Superhuman AI

Find out why 1M+ professionals read Superhuman AI daily.

In 2 years you will be working for AI

Or an AI will be working for you

Here's how you can future-proof yourself:

1. Join the Superhuman AI newsletter – read by 1M+ people at top companies

2. Master AI tools, tutorials, and news in just 3 minutes a day

3. Become 10X more productive using AI

Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.

Sign up and start learning AI

Meta

🧠 Meta Llama AI Just Got Smarter, Chattier—and More Secure

Meta just threw its first-ever LlamaCon, and it was packed with AI firepower.
Top of the list? A new standalone app for Meta AI and a sneak peek at its Llama 4-powered API.

📱 New Meta AI App

• Powered by Llama 4

• Learns your preferences (with permission)

• Offers personalized, context-aware replies

• Supports voice + text, image generation

• Includes a fun “Discover” feed for prompts

🧑‍💻 Llama API Preview (Free!)

• Devs get early access to:

  • Llama 4 Scout

  • Llama 4 Maverick

• Build smarter, faster AI tools

🛡️ AI Safety Upgrades

• 🔐 Llama Guard 4

• 🔥 LlamaFirewall

• 🧪 Defenders Program

  • Early access to security testing tools

🎙️ Bonus: Zuck Speaks Out

On the Dwarkesh Podcast, Mark Zuckerberg talked:

• AGI goals 🤖

• China’s AI race 🇨🇳

• Why open source still matters 🧬

💥 Why It Matters

• 📈 1B+ downloads of Llama already

• Deep integration with Meta’s apps = unmatched personalization

• New tools = more power for creators & devs

📢 AI that knows you. Tools that empower you. Welcome to the next era of Meta AI.

Trending Today

• Clone Robotics has unveiled the first robot with synthetic muscles calling it as Protoclone, that reportedly has over 1000 artificial muscles built in with the company's Myofiber. This makes it possible to flex and walk properly. A 3D printer is used to make parts with polymer and Myofiber muscles. Fused joints reinforce stress points gives Protoclone a smooth movement and flowing motion.

• OpenAI is rolling out a shopping-friendly search mode in ChatGPT, letting users compare prices, read reviews, and buy products directly in the chat. Results include images, pricing, and purchase links—no ads or paid placements. The feature will soon be available to all users, including free tiers.

• Duolingo is replacing contractors with AI for tasks like content creation. CEO Luis von Ahn says AI will handle repetitive work, while humans focus on creative problem-solving. New hires? Only if AI can’t do the job. AI is also being used in hiring and performance reviews.

• Tavus just dropped Hummingbird-0, a zero-shot lip-sync model that matches lip movements to any voice—no training needed. Built from their Phoenix-3 model, it produces hyper-realistic sync from just one video input, outperforming all competitors.

Recommended Reading
If WE had to recommend other newsletters

AI Tool Report is an Newsletter, we enjoy reading daily. AI Tool Report delivers top techniques on how AI can transform your business filled with practical tips, real-world examples.

A Smart Bear is for people who love thinking about strategy, startups, product, marketing, decision-making, and founder psychology.

Job Opportunities

• Sanctuary AI - Product Manager, Physical AI and Software - Apply

• Hebbia - Engagement Operations - Apply

• Dataiku - Associate Data Scientist - Apply

• Mistral AI - Senior Backend Engineer, Product - Apply

• Shield AI - Staff Mission Systems Engineer - Apply

AI News

Ideas? Comments? Complaints?

We read your emails, comments and poll replies daily.

Get the most important AI, tech, and science news in a free daily email.
New Here? Subscribe!
Sponsorship Slots Open for May and Reach over 800+ active readers. (Now 40% off) 🤯

Until next time, Stay Informed!