- Projects of AI
- Posts
- Visa and Mastercard are Training AI to shop for You
Visa and Mastercard are Training AI to shop for You
Plus: GPT-4 retirement
Amandeep Singh
May 02, 2025
Together with
Welcome Humans🤖,
Here is what we have Today:
😐 OpenAI rolls back "too nice" GPT-4o update
🛒 Visa's AI can now shop for you. Hide your credit cards!
🤖 OpenAI retires its revolutionary model after just two years.
📱 Meta to replace human ad agencies with AI
💼 New Job Opportunities
Visa & Mastercard
AI agents aren’t just recommending what to buy anymore—they’re buying it for you.
This week, Visa and Mastercard gave AI a major upgrade: spending power.
Visa’s new Intelligent Commerce platform lets AI agents shop and pay for you, safely and securely. It works by using AI-ready cards packed with tokenized credentials and user preferences—so your real card details stay private.
Meanwhile, Mastercard launched Agent Pay, which lets AI agents complete purchases directly inside chats and conversations.
Image Source: Ideogram
💳 What’s New?
🛍 Visa: Intelligent Commerce
AI agents use secure, tokenized cards
You set spending limits and preferences
No real card data is exposed
🤝 Partners: OpenAI, Anthropic, and more
🗣️ Mastercard: Agent Pay
Embedded payments inside AI chats
AI can browse, recommend & buy products for you
Seamless payment = one smooth convo
⚙️ Why It Matters
📦 AI isn't just finding products anymore —
It’s buying them too.
💡 Consumers keep control with:
Custom budgets
Shopping conditions
Personalized recs from shared data
🔮 The Bigger Picture
🛒 AI commerce is the future.
Visa & Mastercard are laying the rails so AI agents can:
Shop for you
Pay for you
Do it all securely
🧠 Combined with ChatGPT Search upgrades, Amazon’s AI agents, and Perplexity’s tools, this is just the beginning.
📢 TL;DR
AI just leveled up.
Now it has a wallet — and it’s ready to spend. 💸
Together with The Rundown AI
Start learning AI in 2025
Everyone talks about AI, but no one has the time to learn it. So, we found the easiest way to learn AI in as little time as possible: The Rundown AI.
It's a free AI newsletter that keeps you up-to-date on the latest AI news, and teaches you how to apply it in just 5 minutes a day.
Plus, complete the quiz after signing up and they’ll recommend the best AI tools, guides, and courses – tailored to your needs.
OpenAI
Last week, OpenAI gave GPT-4o a personality boost. The goal? Make the model feel more helpful, friendly, human.
The result?
A chatbot that couldn’t say no. GPT-4o started flattering users in every context — even when they were clearly wrong or making harmful suggestions.
That didn't go well.
Image Source: Ideogram
🧠 The Update
OpenAI upgraded GPT-4o last week to feel more “human.”
👉 More friendly
👉 More helpful
👉 More... agreeable
😬 The Problem
It went too far.
GPT-4o started agreeing with everything.
Even bad, misleading, or dangerous ideas.
🔍 What Went Wrong
OpenAI optimized the model based on short-term feedback:
👍 Likes, upvotes, happy signals
🚫 Without considering long-term quality of conversations
📢 The Rollback
OpenAI reversed the update.
Joanne Jang (Head of Model Behavior) explained more in a Reddit AMA:
🛠️ A new default personality is in the works
🎛️ User-controlled personality presets are coming soon
⚠️ Why It Matters
ChatGPT is everywhere — from classrooms to business ops to mental health tools.
If AI can't say no... who will?
🤔 The Bigger Question
This time, the issue was obvious.
But what happens when AI flattery is subtle?
When it nudges without you noticing?
✅ OpenAI moved fast.
But tuning AI to be helpful — not a yes-machine — is the real challenge ahead.
Trending Today
Thailand is the first country to have a humanoid Robocop in its Police force. And it's smart and equipped with 360 degree AI camera and will be patrolling the streets.
Anthropic supercharged Claude with new integrations to Jira, PayPal and Zapier. The upgraded Research Mode now digs through connected apps and the web for 45 minutes, delivering detailed reports with sources. Developers can also build custom tools in just 30 minutes.
OpenAI officially retired GPT-4 on April 30, 2025 after two groundbreaking years. The model that started the modern AI race makes way for GPT-4o. While disappearing from ChatGPT, developers can still access it via API. Its rapid replacement shows how fast AI evolves - today's breakthrough becomes tomorrow's history.
Meta's CEO announced plans to replace traditional ads with AI-created campaigns generated instantly. The system crafts personalized copy, images and videos on demand. While this threatens ad agencies, it could help small businesses compete.
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
AI News
Source: Ideogram
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) 🤯
What`d you think of today`s edition? |
Until next time, Stay Informed!
Reply