Today in AI

1.  Sam Altman walks back AI job-loss warnings ahead of IPO

OpenAI's CEO now says he was "pretty wrong" about AI's near-term impact on jobs, which is a notable reversal from his earlier predictions. The timing matters: OpenAI is reportedly eyeing a Q4 2026 public listing at a roughly $1T valuation. Whether this is genuine reassessment or IPO optics is your call. Read Time's analysis.

2.  A startup just bet $15M that specialized AI beats frontier models

Trajectory emerged from stealth with a platform that continuously post-trains agentic models using real-world product usage. The team pulled researchers from DeepMind, OpenAI, Apple, and Meta Superintelligence. The core idea: models that keep learning from actual use will outperform static frontier models. See how Trajectory works.

3.  Samsung workers locked in $370K AI bonuses after threatening a strike

About 78,000 Samsung workers approved a new labor deal tied to the company's operating profit, which surged roughly 750% in Q1 thanks to AI memory demand. AI is reshaping paychecks, not just job titles.

Join 2M+ Professionals Getting Ahead on AI

Keeping up with AI shouldn't feel like a second job.

But between the new tools, viral posts, and endless hot takes, most people spend hours every week trying to figure out what actually matters.

The Rundown AI fixes that.

It's a free newsletter that gives you the AI news, tools, and tutorials you actually need to know. All in just 5 minutes a day.

Over 2M professionals at companies like Apple, Google, and NASA already read it every morning to stay ahead.

Plus, if you complete the quiz after signing up, they'll recommend the best tools, guides, and courses for your specific job and needs.

From the Frontier

AI in the kitchen

London robotics startup Kaikaku just launched Epicure, a food AI model trained on 4M+ recipes and 1,790 ingredients across 7 languages. One researcher called it "all of human cooking compressed into two megabytes." That's a bold claim. But the early demos are hard to argue with.

How it actually works

Epicure is built on FlavorGraph, a dataset that maps the chemical relationships between ingredients. So instead of pulling from recipe blogs, it generates science-backed meal recommendations based on proven flavor pairings. The result: recommendations that are likely tastier than what any general-purpose LLM would suggest.

Why general models fall short

When Claude or ChatGPT suggests a meal, it's basically educated pattern-matching from training data. Epicure was purpose-built for food. The difference shows.

The bigger trend

Epicure is part of a growing class of vertical AI models: Harvey for legal work, BloombergGPT for finance, AlphaFold for protein structures. The bet is consistent across all of them. The next era of AI will reward depth, not breadth.

Video Prompting: Google published its Ultimate Video Prompting Guide for Gemini. Five core strategies, with real prompts included. Worth bookmarking.

Viral Prompts: These nine general-use AI prompts have crossed 1M views for a reason. They turn your chatbot into a project manager. Give them a look.

Ferrari Redesign: A Redditor who was underwhelmed by the Luce used AI to redesign it. The result arguably beats the original. 2K upvotes don't lie.

Omni Footage: Gemini Omni turned a rough sketch into a hyper-realistic POV drone video. 1M views. Watch it once and you'll understand why people are paying attention.

Claude Code Security: Anthropic's newest built-in plugin for Claude Code spots and fixes vulnerabilities while you work. 1.5M views. If you write any code, this one's for you.

Prompt Station

Cover letter that doesn't sound like every other cover letter

Copy, fill in the brackets, and paste into ChatGPT or Claude. That's it.

Act as an expert career coach, executive resume writer, and persuasive copywriter. Write a compelling, highly personalized cover letter for the role of [JOB TITLE] at [COMPANY].

Use the job description, company mission, values, recent priorities, and role requirements to clearly connect my background to what the company is looking for.

Start with a strong, memorable opening that avoids generic phrases like "I am excited to apply." Use concise storytelling to highlight my most relevant achievements, skills, and career motivations, while showing why I am a strong fit for this specific role and company.

Keep the tone professional, confident, warm, and human. No vague claims, filler, or overused corporate language.

Here is my profile: [DETAILS]. Here is the job description: [JD].

Why it works: Most cover letters fail because they're generic. This prompt forces the AI to tie your specific background to the company's specific needs. It also bans the phrases that make hiring managers skip to the next application.

How 2M+ Professionals Stay Ahead on AI

What’s the secret to staying ahead of the curve in the world of AI? Information. 

Luckily, you can join 2,000,000+ early adopters reading The Rundown AI — the free newsletter that makes you smarter on AI with just a 5-minute read per day.

Keep Reading