OpenAI finally turns on ads

OpenAI crosses the ads line

This was always coming.

OpenAI is testing ads inside ChatGPT for users on Free and the $8 Go plan in the U.S., while Plus, Pro, Business, and Enterprise users remain ad free. Ads will be labeled, separated from responses, excluded from sensitive topics, and not shown to users under 18, according to the company.

The important detail is not the ads themselves, it’s where they show up and who pays to avoid them. OpenAI is signaling that ChatGPT is no longer just a product, it’s a surface with tiers, targeting rules, and revenue pressure that now has to fund large models at scale.

This creates a clearer split between casual users and power users who treat ChatGPT like infrastructure and will pay to keep it clean. It also hands Anthropic an easy contrast, which explains why those Super Bowl ads landed so sharply on Sunday night.

Once ads are normalized here, what other “clean” AI tools start to feel temporary?

AI capex is starting to hurt

The spending numbers are no longer abstract.

Alphabet, Microsoft, Meta, and Amazon are expected to spend close to $700 billion combined on AI infrastructure in 2026, more than 60% above 2025 levels. That surge is pushing free cash flow down sharply, with Amazon projected to swing to as much as negative $28 billion and Meta facing potential negative FCF in 2027 and 2028.

What matters here is not optimism about AI’s future, it’s how quickly cash cushions are getting thinner as companies front load the build. When firms with over $420 billion in combined cash still talk openly about debt, equity raises, and delayed buybacks, the strain is real and visible on earnings calls.

This raises the bar for proof, especially for cloud and model returns that have to justify data centers, chips, and power bills running day and night. It also tightens the link between OpenAI’s trajectory and the broader market, which makes any stumble feel louder.

How patient will investors stay once the lights stay on but the cash keeps draining?

Claude Code is creeping into real workflows

This list feels excessive at first.

The CTO of Dropbox, the creator of Node.js, and Andrej Karpathy are all publicly using Claude Code as part of their daily engineering work. Around that gravity, a loose ecosystem has formed: 160K engineers on one daily newsletter, multiple paid courses, GitHub repos, and YouTube workflows pitching Claude Code as the core tool.

What stands out is not the hype, it’s the repetition across serious builders who ship real systems and quietly standardize on the same assistant. When experienced engineers converge on a tool this fast, it usually means the friction is low and the payoff shows up during long coding sessions at a dim desk late at night.

This starts to shift the baseline skill set, where knowing how to direct an agent matters as much as writing clean functions. Engineers who ignore this may still code well, but they will move slower on complex projects.

If this many seniors are adjusting their habits, how long before this stops being optional?

Prohuman team

Free Guides

Explore our free guides and products to get into AI and master it.

All of them are free to access and would stay free for you.

Feeling generous?

You know someone who loves breakthroughs as much as you do.

Share The Prohuman it’s how smart people stay one update ahead.