AI killing jobs or growing jobs?

Happy Tuesday Brex Community,

Is AI Actually Killing Jobs? I Honestly Can’t Tell. I’ve been going back and forth on this lately, and the more I pay attention, the less clear it becomes.

On one side, you have a steady drumbeat from founders and people building frontier AI models saying job disruption is coming fast. The tone ranges from cautionary to outright alarming, and it’s hard to ignore. If you only listened to that camp, you’d assume we’re on the verge of a major wave of job loss.

But then you look at what’s actually happening inside companies, and the picture feels different. Many of the same organizations going all-in on AI are still hiring, sometimes aggressively. I saw this highlighted recently by Marc Benioff, heavy AI adoption alongside continued workforce growth and it lines up with what I’m seeing firsthand.

At Brex, we’re using AI across the business: customer support, underwriting, risk management, engineering and sales. It’s not theoretical; it’s embedded in real workflows and getting more capable by the month. And yet, we’re also hiring. So my direct experience doesn’t match the “AI is killing jobs” narrative, at least not right now.

That’s where the disconnect sits. You have credible voices warning about what’s coming, and real-world data points suggesting something more incremental and less dramatic in the present.

Maybe it’s a timing issue. It’s possible we’re still in the augmentation phase, where AI makes people more productive but doesn’t yet reduce headcount. Or maybe the impact will vary widely depending on the role, with some jobs changing quickly while others evolve more slowly.

Either way, I don’t think there’s a clean answer yet. The reality feels messy, uneven, and still unfolding. For now, I’m less interested in picking a side and more interested in watching closely. Because whatever happens next, it’s probably going to be more nuanced than either extreme suggests.

Have a great week!

Shai Goldman

Events you might see us at - subscribe here

April 27 - NYC - Applied AI in Fintech - Apply Here

April 28 - NYC - Brex Flex Club - for startup founders - Apply Here

April 29 - NYC - Brex Supper Club - for Pre Partner VCs - Apply Here

May 2 - SF - Brex Flex Club - recover edition - Apply Here

May 5 - NYC - Brex Social Club for GTM (AEs/SDRs) - Apply Here

May 13 - SF - Brex Flex Club - for finance leaders w/ Superpower - Apply Here

May 13 - SF - Brex Flex Club - Padel - Apply Here

May 19 - NYC - Brex Social Club - Apply Here

May 21 - SF - Fireside with Sam Blond of Monaco - Apply Here

Recurring - NYC - Founder workout sessions - Apply Here

Meet Kepler and it’s Founders, Vinoo and John.

The Problem
The bottleneck for enterprise AI isn’t intelligence, it’s trust. After 150+ conversations with hedge funds, asset managers, investment banks and PE firms, Kepler’s ex-Palantir founders kept hearing the same line: “I can see how powerful this is, I just can’t trust it.” Analysts can’t show their work, can’t trace a number to its source, and can’t defend an AI output in front of an investment committee or a regulator - so they keep doing the work manually. The infrastructure supporting today’s AI tools was never designed to be accurate or traceable.

The Solution
Kepler is deterministic infrastructure built for AI. Kepler separates reasoning from retrieval: the LLM interprets intent, then deterministic code retrieves the data, runs the calculations and cites every source. The model never generates a number, never touches a calculation, and never decides what’s true. That separation makes every output traceable to its source, auditable step by step, and defensible under any review. Their first product, Kepler Finance, is live with partners querying across 26M+ SEC filings, 14,000+ companies, and 27 global markets.

Why It Matters
Every industry where being wrong is expensive needs a way to prove AI is right, and that gap is wide open. Kepler’s design partners already span hedge funds, investment banks, university endowments, and Fortune 1000 industrials. They are backed by investors who built the modern AI and data stack including founders of OpenAI, Meta AI Research, MotherDuck, and dbt Labs. Their product is live, tested and proven - now it’s about deployment and scale.

Haystack raised a new $85M fund.

Haystack is a specialized early-stage venture capital firm that focuses on the foundational chapters of software and technology startup formation. For over a decade, the firm has positioned itself as a partner for outlier founders, typically investing between $1M and $3M in initial funding rounds as either a lead or participating investor. Their portfolio is distinguished by its early backing of industry giants such as DoorDash, Instacart, Figma, and HashiCorp, demonstrating a strong track record in identifying transformative platforms across diverse sectors including AI, infrastructure, and consumer services. Operating as a lean investment team, Haystack emphasizes a deep network of operators and industry experts to support their founders from the earliest stages of development.

Sign up here for Brex and receive $1,250 towards Brex Travel or $750 cash

Brex is the intelligent finance platform built for speed and control. Combining corporate cards, banking, treasury, expense management, bill pay, travel, and accounting, Brex helps companies spend smarter and move faster in 120 countries. By integrating AI across every workflow, Brex enables founders and finance teams to eliminate manual work, do more with less, and accelerate impact. More than 35,000 companies from startups to enterprises run on Brex, including Anthropic, Cursor, ServiceTitan, Robinhood, DoorDash, and Sonos