OpenAI CEO says AI won't cause a "jobs apocalypse"

Welcome back, Superhuman. Sam Altman no longer thinks AI will cause a "jobs apocalypse." The OpenAI CEO now argues the "human part" of employment can't be replaced — a notable reversal from his earlier warnings. Time has some ideas on what changed his mind.

Today: Sam Altman takes back AI job loss statements, how to optimize your LinkedIn profile with Gemini, and get the latest prompts and trending social posts.

1. Sam Altman walks back AI job-loss warnings ahead of IPO: In a recent interview, OpenAI CEO Altman said it’s unlikely that AI will create a jobs apocalypse and that he was "pretty wrong" about AI's near-term impact on the labor market. This comes as OpenAI is reportedly preparing for a Q4 2026 listing, aiming for a roughly $1T valuation. Read Time’s analysis.

2. Trajectory emerges from stealth betting on “Continual Learning”: Trajectory is building a platform that uses real-world product usage to continuously post-train agentic models, a strategy it's betting can outperform frontier models. The startup has raised $15M and has pulled in researchers from DeepMind, OpenAI, Apple, and Meta Superintelligence. See how Trajectory works.

3. Samsung workers lock in $370K AI bonuses after threatening strike: Samsung Electronics' largest union approved a new labor deal that makes roughly 78,000 workers eligible for annual bonuses of about $370,000. The bonuses are tied to the company’s soaring operating profit, which was up roughly 750% in Q1 due to AI memory demand.

No, it’s not your model or your team—it's your data.

The strongest agentic AI in the world falls apart without the right data foundation underneath it. So how do you build one that actually scales?

Check out this panel from AWS. It gathers execs from Yahoo, Mercedes-Benz, and Regeneron, unveiling how they:

Register here (costs nothing).

AI in the kitchen. London robotics startup Kaikaku just dropped Epicure, a food AI model trained on 4M+ recipes and 1790 ingredients across 7 languages. Josef Chen calls it, "all of human cooking compressed into two megabytes” — and it could help answer the age-old question: What’s for dinner?

Pairing professional. Epicure is built on FlavorGraph, a dataset of food ingredients and chemical compounds that maps how flavors interact. This allows Epicure to generate science-backed meal recommendations based on chemically-proven flavor pairings. In other words, Epicure’s recs are probably much tastier than what Claude, ChatGPT, or another LLM would suggest.

Why general models fall short. When Claude or ChatGPT makes a meal recommendation, it’s really just an educated guess — pulled from whatever recipe blogs or Reddit comments ended up in their training data. Epicure was built from the ground up for food recommendations.

The vertical trend. Epicure joins a growing class of vertical AI models built for deep domain expertise rather than broad utility. Others include Harvey for legal work, BloombergGPT for finance, and AlphaFold for protein structures. The bet underneath all of them is the same: the next era of AI will belong to specialized models, not generalists.

Sample Prompt: [upload profile] Before rebuilding and optimizing my LinkedIn profile, ask me a few clarifying questions. Request my primary goal (sales calls, leads, job offers, newsletter growth, authority positioning, etc.). Ask for my target audience (be specific). Ask for my main offer (what I want people to buy, book, or join). Ask for up to two Featured links (URL + description). Ask for key social proof to highlight (audience size, revenue, brands, media mentions, etc.). Once you have this, rewrite my full LinkedIn profile including headline, About section, and experience.

That’s how you optimize your LinkedIn profile professionally.

Capture live traffic, fine-tune and optimize, then deploy your own checkpoints to dedicated GPU endpoints.

Choose hardware, set scaling limits, and select region. Stable latency, predictable cost, clear data residency.

From LLM to production system, in one platform

🎥 Video Prompting: Master Gemini Omni with Google’s Ultimate Video Prompting Guide that includes 5 core strategies, with plenty of prompts included.

📝 Viral Prompts: Upgrade your chatbot from a Q&A machine into a full-blown project manager with these nine general-use AI prompts that have racked up 1M+ views.

🏎 Ferrari Upgrade: This AI-generated redesign of the Ferrari Luce looks arguably better than the real thing, created by a Redditor who was underwhelmed with the Luce’s design (2K upvotes).

🤯 Omni Footage: Gemini Omni converted a rough sketch into an insane POV-style drone video that’s gone viral for its accuracy and realism (1M views).

🔒 Claude Code: Anthropic’s newest built-in security plugin spots and fixes code vulnerabilities while you’re working, keeping your projects secure (1.5M views).

^{* indicates a promoted tool, if any}

The internet agrees that this phrase has replaced the em dash as the most obvious sign of AI-generated text — try to avoid using it in your work whenever possible (7M views).

Grow customers & revenue: Join companies like Amazon, HubSpot, and Salesforce. Showcase your product to our 1M+ readers and 2M+ followers on socials. Get in touch.

Until next time — Zain, Theodore, & the Superhuman AI team.