Major AI Gains Are Shifting Pay, Power, and Priorities

Pivot 5
June 04, 2025

Pivot 5 is the must-read daily AI briefing for 500,000+ CEOs and business leaders who need signal, not noise.

5 headlines. 5 minutes. 5 days a week.

PwC Says AI-Exposed Jobs Are Growing Faster and Paying More

PwC released its 2025 Global AI Jobs Barometer after analyzing almost one billion job ads across six continents. The report finds that AI-exposed roles are increasing in number and earning higher wages rather than disappearing.
Since generative AI took off in 2022, productivity in heavily exposed sectors like financial services jumped from 7% to 27%, and those industries now post revenue per employee that is three times higher than low-exposure fields. Roles requiring AI skills carry an average 56% wage premium and grew 7.5% over the past year even as total job postings fell 11.3%.
Job counts are rising in every AI-exposed occupation, including highly automatable ones. Employer demands are shifting quickly, with desired skills changing 66% faster and degree requirements dropping notably in AI-related jobs.

Read more here.

Snowflake announced it will acquire Postgres specialist Crunchy Data to launch an enterprise-grade Snowflake Postgres service. A source familiar with the transaction puts the price near $250 million, a figure Snowflake declined to confirm.
Crunchy Data builds tooling for the open-source PostgreSQL database and serves clients including UPS, SAS, Moneytree, and the U.S. Department of Homeland Security. Snowflake plans to offer the technology in private preview soon, expanding on connectors it introduced last year for PostgreSQL and MySQL.
The deal follows recent data-infrastructure buys by Salesforce, Alation, and ServiceNow aimed at powering AI agents. Snowflake says the move targets a $350 billion market and strengthens its claim to be the “ultimate destination” for enterprise data and AI workloads.

Read more here.

IBM acquires New York–based data analysis startup Seek AI for an undisclosed amount and unveils the WatsonX AI Labs accelerator in Manhattan. Seek’s natural-language query engine becomes a core component of WatsonX to let enterprise users interrogate data through chat-style prompts.
Seek, founded in 2021 and backed by roughly $10 million from Battery Ventures and others, converts plain-language questions into database queries and summaries. The startup will move its headquarters to One Madison, the site of WatsonX AI Labs and IBM's new Manhattan offices, once the deal closes.
IBM says doubling down on enterprise AI has already lifted its Q1 software revenue and beaten earnings estimates. Locating WatsonX AI Labs in Manhattan taps the city’s diverse talent pool and facilitates collaborations with local colleges and research institutions.

Read more here.

The article maps a widening split between AI leaders betting on artificial general intelligence within five years and experts who see decades of work ahead. It structures the dispute around five arguments for short timelines and four reasons progress may stall.
Optimists point to frontier models topping 82% on MMLU, task-duration capacity doubling every seven months, and CEOs like Dario Amodei and Sam Altman expressing confidence in sub-decade breakthroughs. Skeptics counter with surveys that still place median AGI arrival after 2047, lingering robot and real-world task failures, and doubts that an “intelligence explosion” can beat compute and data limits.
The author highlights 2030 as a practical cutoff: if AGI is absent by then, trillion-dollar hardware costs could curb further scaling and push timelines out. She underscores that the world is “drastically underprepared” for the near-term AGI outcome many insiders now consider plausible.

Read more here.

Early-stage GTM advisor Alex Shartsis built a full-featured ROI calculator in under two hours using Bolt.new and other “vibecoding” tools. The interactive React app replaced a dense spreadsheet, captured prospect emails, visualized valuation gains, and ran live without engineering help.
Shartsis supplied Bolt with screenshots of the Excel model and a prompt detailing inputs, branding, and step-by-step UX, then let the agent generate and host the code. Pico’s founder said an agency had quoted about $10,000 and a month for comparable work, underscoring the tool’s speed and cost edge.
The article shows how AI coding frees GTM teams to launch calculators, scrapers, forms, and landing pages for pocket change and minimal skill. That autonomy turns once cost-prohibitive experiments into quick assets that directly drive pipeline.

Read more here.