You've probably seen the headlines about Google, OpenAI, and Anthropic all racing to build the smartest AI.

Turns out the race isn't just about the technology. It's about who can keep their best people from walking out the door.

This month, Google lost two of its most important AI scientists. One went to OpenAI. One went to Anthropic.

We break down why that matters even if you don't work in AI, plus fresh data on who's actually hiring right now, where the money's flowing into new companies, and a copy-paste prompt for your next raise conversation.

Let's get into it.

Funding

Contracts

Leadership is asking: are we getting value from AI? Which tools are worth the spend? Where are we exposed? Right now, most teams have no idea.

Harmonic Security Usage Explorer changes that.

You get a complete picture of how your organization uses AI, automatically categorized into custom tasks and use cases.

You’ll see the projects being worked on, who’s using what tools, where AI investments are driving value, and where employees are engaging in risky behavior.

CIOs can rationalize spending and cut wasted licenses. CISOs can pinpoint where risk exists and neutralize it. AI committees can show exactly how their efforts are paying off.

See it in action

Google's DeepMind lab, the team behind its most advanced AI models, lost two of its most important people this month, and they didn't even go to the same place.

Noam Shazeer, one of the researchers who co-invented the "Transformer," the architecture nearly every major AI model (ChatGPT, Claude, Gemini) is built on, left for OpenAI. Around the same time, John Jumper, who led the team behind AlphaFold (the AI that predicts how proteins fold, a genuine breakthrough for drug discovery), left for Anthropic.

Two different people. Two different destinations. Same lab, same month.

This isn't really a story about Google messing something up. It's a story about scarcity. There are only a few hundred researchers on the planet who've actually built a frontier AI model from scratch, and every major lab, OpenAI, Anthropic, Meta, xAI, wants the same short list of names.

That scarcity is why pay for this specific group of people stopped looking like a normal job offer and started looking more like pro sports free agency: multi-year guarantees, equity packages worth hundreds of millions, sometimes entire teams poached together instead of one person at a time.

It also explains Google's other move this month: reorganizing its AI leadership back to California and simplifying who reports to whom. That's not symbolism, it's a retention play, an attempt to make the org fast and flat enough that the next Shazeer or Jumper has fewer reasons to leave.

Why it matters: You probably don't work at a frontier AI lab, but this still tells you something about your own job. When a market is this scarce at the very top, it pulls pay and demand up through the whole field underneath it, from AI researchers down to anyone who can competently deploy AI tools inside a normal company. If your job touches AI at all, even loosely, this is a good week to check what the market actually pays for what you do, because the number has probably moved since you last looked.

Wild to think a couple of people switching jobs can move an entire industry's pay scale. -San

Three years ago, AI-related certifications made up maybe 1 to 2% of all the professional certificates people were earning. Today they're close to 30%, something one analysis called a 20x jump from pre-ChatGPT levels.

Employers are asking for it too. Entry-level job postings that specifically require AI skills nearly doubled in about six months, from 10.5% of postings last fall to 16.5% this spring. More than a third of employers now say entry-level hires need to show AI fluency before they'll even get an interview.

Here's the part most people miss. A certificate only moves your pay if it's actually related to your job. Workers whose first non-degree credential matched their real field saw a 3.8% wage bump. Workers who got a random, trending AI certificate that had nothing to do with their job saw only 1.8%, barely more than nothing.

The same "be specific, get paid" pattern shows up in job-switching too. People who changed jobs in July got paid 7% more than they were making a year earlier, the widest gap in almost a year, concentrated almost entirely in healthcare and AI-related roles. Switch into a generic role instead, and that premium mostly disappears.

Why it matters: Don't collect AI certificates like trophies. Pick the one specific skill gap standing between you and the job you actually want, get certified in that, and put real examples of you using it right next to it on your resume. That's the version of "AI skills" that actually shows up in paychecks.

If your AI cert has nothing to do with your job, it's decoration, not a raise. -San

Land your next role faster with Tailr, our Claude-powered job search and networking tool: Start searching

The Raise Ammunition Prompt

Copy this into ChatGPT or Claude, free tier is fine:

"I want to build a data-backed case for a raise. Here's my situation: [your role, current pay, time in role, and 2-3 concrete wins from the last 6-12 months]. Do three things: (1) Estimate a realistic market range for my role using public salary data, and tell me plainly if you're not confident in a specific number rather than guessing. (2) Tell me which 2 of my listed wins are the strongest to lead with, and why. (3) Write a 100-120 word message I could send my manager that references the market data (national wage growth is running about 3.2% right now, but people who switch jobs are seeing closer to 7%) without sounding like I'm threatening to quit."

Last issue we asked whether July's inflation report would come in hot, in-line, or cool. It came in cooler than expected, with core inflation hitting its slowest pace since 2021.