Most Data Professionals Are Preparing for the Wrong Future

Executive summary (business-focused)

• Shahar argues the “AI will take your job” narrative is incomplete. Layoffs and cost-cutting align strongly with macroeconomics (notably interest rate hikes), while AI mainly changes how work is done: shifting from “instructing” systems to “requesting outcomes.”
• Data professionals remain essential, but the definition of “value” shifts. The winning advantage moves from technical production to clarity, direction, leadership, ownership, and business understanding—because AI makes it easy for many teams to build things, including the wrong things.
• Career and organizational success in the next era depends less on “ability to do work” and more on ability to do the right work, plus personal differentiation via networks, personal brand, creativity, frameworks/playbooks, and mobility.

Industry “crisis” timeline & operating reality

Observed pattern (business/ops context):

• 2018–2019: “normal” hiring and growth
• 2020: COVID shock → rapid tech growth, hiring surges (e.g., Meta recruiting targets increased)
• ~Q2 2022: shift to layoffs after a long hiring period; increased emphasis on efficiency/cost cutting
• Nov 2022 onward: ChatGPT → AI narrative accelerates
• 2023: cloud/compute expansion + “AI arms race” (companies replacing humans with GPUs; hoarding compute)

Key claim: AI is a huge shift, but the primary execution problem at many companies is leadership/strategy (reinvention, operating model), not simply “AI replaced people.”

High-level framework: what’s changing vs. what’s not

1) What changes (AI-enabled interface + reduced barriers)

• From “how-to” to “what-to-do”: earlier eras required detailed programming instructions (assembly → C → C++/Java → Python), while modern AI enables outcome-based requests.
• Barriers collapse: more people can build and ask questions without traditional skills (e.g., SQL/Tableau users can generate outputs via natural language).
• Risk: lowering barriers also lowers quality control—teams can produce “wrong” answers and “wrong” analyses faster.

2) What doesn’t change (value creation stays central)

Even when automation replaces execution (examples given: autopilot, smartphones for photography, automated flight booking), people still exist to:

• manage decisions, edge cases, exceptions, and premium workflows
• create value in a redesigned capacity

“Winning” operating traits / people playbook (leadership & execution)

Shahar lists traits that function like a capability model for individuals and teams:

• Adaptability & curiosity
  • continuously learn tools/frameworks; what got you here won’t get you there
• Clarity
  • turn noise into signal; communicate complex ideas simply
  • align teams on the right questions/metrics/taxonomies
• Influence / leadership
  • build shared mission and direction amid chaos; lead people through change
• Ownership
  • don’t only surface problems (“funnel is broken”)—define solutions and drive action with other teams
• Business understanding
  • challenge business requests, translate intent into value, and prevent “order-taking”
  • example: instead of building a report the business asked for, ask what goal they’re trying to achieve (e.g., “optimize the marketing funnel”) and steer toward the correct solution

“Direction over speed” (governance becomes harder)

• With AI tools and conversational interfaces, organizations may lose the governance they built (alignment on metrics, taxonomies, and question frameworks).
• Resulting principle:
  • Direction becomes more important than how fast teams can execute.
• The key skill is knowing what questions to ask, not only building.

Employment-era strategy: career assets in a crowded market

Shahar reframes employability as “value + differentiation”, not simply “ability to do tasks.”

Core thesis

• AI makes execution easier for many; the differentiator becomes:
  • right work
  • leadership/ownership/business framing
  • career “assets” that compound over time

Career differentiation assets (actionable)

• Network: who will help you quickly (warm intros, direct collaboration)
• Personal brand: who knows you, how you’re positioned, what you stand for
• Creativity: AI is “repetition”; creativity creates advantage
• Frameworks & playbooks: reusable methods you can apply across contexts
• Location/mobility: ability to access key hubs/opportunities

Concrete examples & analogies (used as business lessons)

• Autopilot: pilots still critical, but shifted to decision-making and crisis management.
• Smartphones vs professional photographers: broad enablement; pros remain for premium/specialized needs.
• Animation automation: computers replaced frame-by-frame labor, but the industry grew and roles changed.
• IBM + flight booking (American Airlines):
  • contact center previously ~600–700 people
  • average booking time ~90 minutes
  • ~8% error rate
  • modern mobile booking achieved ~5 minutes booking and near-zero error (claimed)
  • people still exist for edge cases/cancellations.
• Luddites (textile): job loss fears were real, but outcomes favored new industry scale; workers persist in different roles/capacity.

Recruiting/process recommendations (execution tactics)

• Avoid “gaming ATS with keywords”:
  • Shahar criticizes keyword/AI CV stuffing and suggests it’s a losing strategy in a noisy market.
• Use the “back door”: networking + in-person discovery
  • attend meetups, meet people, and build relationships directly
• Personal content strategy
  • post and share insights, but avoid “AI slop” content automation
• Graduates / job-seekers advice
  • do real projects that solve real problems for businesses (not Kaggle-only score-chasing)
  • talk in interviews about outcomes achieved for actual stakeholders

(No formal recruiting framework named, but the recommendations form an implicit playbook: network → demonstrate real business impact → build credibility publicly.)

Metrics / KPIs mentioned (limited, high-level)

• Interest-rate alignment with the layoffs/cost-cutting timeline (no numeric targets provided).
• Block (Twitter/X) example: workforce cut 40% (10,000 → 6,000 employees).
• Meta examples (qualitative):
  • data engineering site growth: ~4 data engineers → 200+ at peak
  • ads portfolio data managed: $7B/year run rate
• Flight booking case (quantitative):
  • booking time: ~90 minutes (call center era) → ~5 minutes (phone era)
  • error rate: ~8% (call center era) → ~0 / zero (modern booking era as stated)

(No explicit data-team KPIs like CAC/LTV/churn were discussed.)

Presenters / sources

• Presenter: Shahar
  • data executive; formerly Meta/London data engineering site lead
  • director of data engineering for Facebook/Instagram/Messenger trust & safety
  • 20+ years experience; also cites adviser/advisory work
• Referenced sources/figures:
  • Mark (Meta CEO): remote work prediction for ~2030 (half workforce remote)
  • Jack Dorsey (Block/Twitter founder): referenced workforce reduction announcement (40%)
  • Patrick McFadin: flight booking example referenced (“in day-to-day Texas in January”)
  • Daniel Priestley: referenced talk on niching/entrepreneurship (“why everybody needs to be an entrepreneur”)
  • Enzo: mentioned as potentially disagreeing with niche vs generalist framing