🧙🏼 Claude Learning is here

Howdy wizards,

Did you see? GPT-5 is coming in a few months. Exciting times!

Also—your hack for this week if you’re a ChatGPT Plus subscriber is to use Sora for generating images. It uses the new image generator, but doesn’t rate limit you—only limits you to 2 generations at a time. Have fun!

Here’s what’s brewing in AI.

DARIO’S PICKS

1. Anthropic launches Claude Learning Mode — the study-buddy version of Claude

Anthropic just launched Claude for Education, a specialized version of Claude designed to develop critical thinking rather than just doing students' homework for them. The key feature is "Learning Mode," which uses Socratic questioning to guide students through problems instead of serving up ready-made answers.

Three major universities have already jumped on board with campus-wide deals: Northeastern (50k users across 13 campuses), London School of Economics, and Champlain College. These partnerships give both students and faculty access to Claude for everything from research papers to administrative tasks.

Claude’s Learning Mode in a nutshell:

• Aims to preserve critical‑thinking skills in a world where AI answers are just a prompt away. For example, it prompts students with “How would you approach…?” instead of dropping the answer.

• Uses Socratic questioning to probe evidence, reasoning, and alternative viewpoints. Surfaces core concepts behind each problem to cement understanding.

• Lives inside Projects, so students keep all chats and resources organized by assignment/topic.

• Has ready‑made templates for research papers, study guides, outlines, etc.

Students can also apply to be Claude Campus Ambassadors or snag API‑credit grants for their own projects—but only after their university signs up (institutional request form here).

‎ Why it matters‎ ‎ Rather than pretending students aren't already using AI for study materials, Anthropic is giving universities a chance to acknowledging and embrace reality, and smarly positioning Claude as a thinking partner.

You might remember the article about AI flashcard generators on my blog from a while back. With Learning Mode, Claude now offers many of the features of these tools natively.

IN PARTNERSHIP WITH WORDWARE

Stop waiting for engineering - build AI agents with no code

Wordware.ai now integrates with 2,000+ apps & data-sources!

Connect your favorite apps, build AI agents, and let any trigger start automated workflows that do actual work for you.

It’s 2025, you’re smart, you know what good looks like - you can just.. build things!

Free monthly credits until we drain our $30M seed round 😂

Try wordware

DARIO’S PICKS

2. AI passes the Turing test 73% of the time, new study finds

The Turing test is a test to see if if a computer can fool a human into thinking it’s human. If it does, it passes.

UC San Diego researchers just showed that GPT-4.5 passes the original three-party Turing test when given a persona prompt. It fooled human judges 73% of the time (that’s well above chance). GPT-4.5, OpenAI’s latest model with greater understanding of emotional intelligence and human intent, was the only one that passed the test. The test also included Meta's LLaMa-3.1, which at 56%, did not pass.

Here’s how they set the experiment up:

• Each participant played judge in four rounds, facing different AI models each time.

• Judge sat at a text‑only web chat with two partners at once: one real person, one AI.

• Conversation ran 5 minutes; judge could message both back and forth.

• After time was up, judge picked which partner was human and rated confidence.

‎ Why it matters‎ ‎ We’re living in absolutely fascinating times — we can no longer tell if things are human, or just math 😄 

UP CLOSE

I turned myself into 24 cartoon styles using ChatGPT’s new image generator

I’ve taken ChatGPT’s brand-spanking-new image generator for another spin (last week: graphic design tasks). This week, a more systematic one, testing it’s key feature, which we all know is: transforming your images into cartoons.

Pick your fighter:

THE PRELUDE TO A WELL CONCEIVED AI STRATEGY

context windows — 760 real-world AI case studies

contextwindows.ai is now a month old. It’s the most complete collection real-world AI implementation case studies anywhere on the web.

• Filter by industry, agent type, AI provider, and more

• See what's actually working in your industry

• Discover which AI strategies are gaining traction 6 months before your competitors even notice them

• Dozens of new case studies from major AI labs added weekly

I’m still running a ridiculous deal for my subscribers to get access for free. Soon, you’ll only be able to get access on shady internet alleys.

⚫️ Refer 6 actual humans to this newsletter (my cult muhaha) and get free lifetime access

The oompa loompas will promptly unlock the content for you at the 6 referral mark. Check your count below.

Explore 700+ AI case studies →

THAT’S ALL FOLKS!

Affiliate disclosure: To cover the cost of my email software and the time I spend writing this newsletter, I sometimes link to products and other newsletters. Please assume these are affiliate links. If you choose to subscribe to a newsletter or buy a product through any of my links then THANK YOU – it will make it possible for me to continue to do this.