Scott Belsky's 12 outlooks for 2026: talent arbitrage, hardware moats, and the SaaS vibe-coding threat

Summary

Scott Belsky outlined 12 predictions for 2026 centered on talent reshuffling, hardware consolidation, and the erosion of bloated SaaS by cheap AI tools that let internal teams build what they need.

Talent arbitrage

Younger workers fluent in AI tools will outpace older colleagues still running analog workflows. Marketing teams still running focus groups instead of AI-powered market research, or hiring SEO experts when answer engine optimization is already better, illustrate the gap. Belsky calls 2026 a "precious work window for younger overlooked talent to gain an advantage as early adopters." The advantage is temporary. Once everyone learns the tools, the gap closes.

Content creation splits

Creators willing to trade control for speed will use AI to pump out ads and social content. Artists unwilling to trade control will adopt tools that preserve precision. Craft-forward technology gets lasting traction; prompt-based slop stays confined to short-form social. At A24, Belsky also predicts that personalized films with audience cameos will lose to people's preference for shared experiences, like a Taylor Swift concert you can discuss with others rather than a one-off film tailored to you alone.

Authenticity becomes scarce

As AI-generated content floods feeds, audiences develop a "membrane of doubt" about authenticity. Fake becomes the default reaction. Attention-grabbing AI launch videos and vibe reels will be ignored unless they contain genuine ideas worth amplifying.

Health data as competitive advantage

Wearables, blood testing, preventative scans, and AI health coaches will give people materially more insight into their own health. Life insurers with de-risked annuity exposure will benefit as insured populations live longer. Quest Diagnostics and biomarker detection platforms will grow. Travel, entertainment, dating apps, and others will shift as longevity becomes mainstream.

Hardware consolidates power

The 2025 surge in hardware startups—AR/VR glasses, racing sims, consumer devices—signals that hardware and operating system coupling will concentrate AI power. As open source models become capable enough to run locally on phones and computers, competition shifts from cloud models to chips, operating systems, and devices. Apple's ability to ship Gemini integration locally matters more than cloud capabilities. Belsky flags Apple's annual release cadence as a constraint when AI models improve weekly.

Proprietary data wins

As connectors between apps become ubiquitous and every company syncs everyone's data, commodity data becomes cheap. New moats are proprietary social graphs (who works with whom), portable personalization for consumers, and real-time data sources captured by robots. Weather, ocean tides, and live data streams become more valuable than static datasets.

Summarization scales and darkens

Tools like Granola already record and summarize Zoom meetings. The implication cuts deeper: granular understanding of what work actually gets done inside companies, who does what, and which roles could be eliminated. Some founders worry that Granola-style summarization could flag redundancies and job displacement opportunities to executives.

Vibe coding eats SaaS

Companies will spawn internal teams to build tailor-made solutions instead of paying for expensive, clunky single-purpose SaaS. The targets are not marquee software like Salesforce or Snowflake, which are growing 5 to 30 percent year-over-year and are not dying soon. The targets are the long tail of overlooked, small-ticket PE-backed software that raises prices without adding features, or mission-critical but outdated tools at mid-market companies. A freelance vibe-coder economy will likely emerge, selling custom apps to business owners the way web developers once did.

The tension is real but bounded. SaaS giants have regulatory moats, proprietary data, and entrenched relationships that simple vibe coding cannot replicate. But software that has no moat, just rent-seeking, is vulnerable.