Welcome back to Daily Zaps, your regularly-scheduled dose of AI news ⚡️ 

Here’s what we got for ya today:

• 📈 AI skills command a 56% wage premium

• 🤖 Microsoft bets on agents

• 🍎 Apple’s foundation model is getting better

• 🗞️ Washington Post plans on using AI to source content

Let’s get right into it!

Sign Up | Advertise | Tools

CAREERS

AI skills command a 56% wage premium

Since the launch of ChatGPT 3.5 in 2022, revenue growth in AI-exposed industries has accelerated sharply—nearly quadrupling—driven by companies rapidly adopting AI to boost productivity and value creation. PwC’s analysis shows these industries experience three times higher revenue per employee growth and wages rising twice as fast compared to less AI-exposed sectors.

Despite concerns, even highly automatable jobs are seeing wage increases, as AI augments rather than devalues roles. The demand for AI-related skills is evolving 66% faster than in other jobs, more than 2.5 times faster than last year, with workers possessing AI skills commanding a 56% wage premium across all industries, highlighting the growing value of AI expertise in the work.

Full 27 page PWC report here.

BIG TECH

Microsoft bets on agents

In March, Microsoft executive Jay Parikh explained how AI technology is becoming cheaper and easier to use because new, more efficient models are emerging. This means many companies will soon be able to build their own AI-powered tools. Microsoft plans to take advantage of this by creating a platform on its Azure cloud where businesses can build and run AI “agents”—automated programs that can handle tasks like managing spreadsheets or fixing website issues with little human help.

Instead of charging customers based on the number of users, Microsoft wants to shift to charging based on how much customers actually use the AI services, which is expected to grow as these agents become more common. Parikh’s team is bringing together different parts of Microsoft’s AI and developer groups to support this new focus, improve reliability, and help customers adopt these AI tools more easily.

You’re stuck in the unpredictable cycle of cold outreach and the occasional referral.

You know organic LinkedIn content should work to drive traffic — but content production feels like a massive uphill battle.

But what if, instead of staring at a blank page trying to "create content," you're already sitting on a goldmine of industry-specific content ideas:

• Video calls with customers

• Internal meeting recordings

• Webinars you’ve hosted

• Interviews, podcasts, and presentations you’ve done

You’re having conversations that are hyper-relevant to your ICP every day. Those transcripts are packed with insights — perfect material for your next LinkedIn post.

All you need is a simple, scalable system that pulls those insights from your existing content and turns them into ready-to-post drafts — with zero extra effort.

Imagine:

• Showing up in the feeds of thousands of ideal prospects every month

• Never needing to carve out hours to brainstorm and write

• Posting consistently with content that attracts prospects and resonates

• Building lasting trust and authority, so you’re top-of-mind when your ideal customer ready to buy

If that sounds good to you, join the Markfive newsletter. We talk about how to build a pipeline full of warm leads that already like and trust you, through AI-enabled content marketing (that isn’t just generic AI slop).

Join the Markfive newsletter →

BIG TECH

Apple’s foundation model is getting better

Apple is actively testing multiple versions of its foundation model, with sizes ranging from 3 billion to 150 billion parameters. The largest model is designed to operate via Apple's Private Cloud Compute infrastructure, which supports its cloud-based AI services.

Despite these advancements, Apple has opted not to release a ChatGPT-like conversational AI. This decision is attributed to concerns over AI-generated "hallucinations"—instances where models produce plausible-sounding but incorrect or nonsensical information. Such issues are prevalent in large language models and have led to philosophical debates within the company about the risks of misinformation and the ethical implications of deploying such technology.

MEDIA

Washington Post plans on using AI to source content

The Washington Post is launching an initiative called "Ripple" to diversify its opinion content by incorporating submissions from external contributors, including Substack writers and nonprofessional authors. Central to this effort is "Ember," an AI writing coach designed to assist contributors in structuring their pieces by offering tools like a "story strength" tracker, guidance on developing a thesis and supporting points, and prompts for crafting a compelling conclusion. 

While Ember will aid in the writing process, human editors will conduct final reviews before publication. The Ripple project is set to begin testing in the fall, with initial partnerships expected to commence in the summer. This development follows recent editorial shifts at the Post, influenced by owner Jeff Bezos's desire to steer the publication toward supporting "personal liberties and free markets," a move that has sparked internal debates and led to changes in the opinion section's direction.