跪拜 Guibai
← All articles
Frontend · JavaScript · Interview

Tech Laid Off 165K in 2026, but ByteDance’s Frontend Hiring Jumped 23%

By kyriewen ·
Read original on juejin.cn ↗ Google Translate ↗ Alt translation

The frontend role is not dying—it is bifurcating. Engineers who only translate designs into code are being replaced by AI at scale, while those who can architect systems, integrate AI APIs, and judge machine-generated output are seeing demand and salaries rise. The shift redefines what 'senior' means: it is no longer about years writing React, but about the density of judgment calls AI cannot make.

Summary

A new dataset from programs.com shows 165,000 global tech layoffs through July 2026, with Oracle alone cutting up to 30,000 roles and Microsoft explicitly citing AI-driven efficiency for its 4,800 cuts. More than 91,000 of those eliminated positions were directly attributed to AI replacement. At the same time, ByteDance opened over 5,000 campus recruitment positions in China, with frontend and client-side roles up 23% year-on-year, and Taotian Group planning roughly 1,000 offers where technical roles exceed 90%.

The contradiction resolves when you read the job descriptions. Traditional frontend roles demanding HTML, CSS, JavaScript, and framework proficiency are closing en masse. The roles still open—at JD.com, ByteDance, and Alibaba—now require AI product integration, TensorFlow experience, streaming rendering, and the ability to handle AI output uncertainty. The keyword shift in JDs is stark: 'proficient in React' is being replaced by 'deep understanding of AI business requirements' and 'AI output verification and quality control.'

A Big 4 consulting report projects that by December 2026, 40% of current jobs will list AI skills as mandatory, not preferred. The frontend role is splitting into two categories: execution-type engineers who write pages from mockups—80% replaceable by AI—and judgment-type engineers who design architectures, integrate AI, and catch the edge cases AI misses. The 165,000 cuts hit the first group; the 23% hiring increase targets the second.

Takeaways
Global tech layoffs hit 165,000 in the first seven months of 2026, with over 91,000 roles explicitly attributed to AI replacement.
ByteDance's 2026 campus frontend hiring grew 23% year-on-year, while Taotian Group plans roughly 1,000 offers with technical roles above 90%.
Traditional frontend JDs requiring HTML/CSS/JS and framework proficiency are closing; open roles now demand AI integration, TensorFlow, NLP, and streaming-rendering experience.
Job descriptions are dropping 'proficient in React/Vue' in favor of 'deep understanding of AI business requirements' and 'AI output verification and quality control.'
A Big 4 consulting report projects 40% of current jobs will require AI skills as mandatory by December 2026, not as a preference.
Five frontend sub-fields are still growing: AI application frontend, data visualization, cross-platform architecture, frontend engineering/DevOps, and full-stack AI engineering.
Execution-type frontend roles—writing pages from mockups—are being replaced by AI at 80% or more; judgment-type roles—architecture, AI integration, quality control—are increasing.
On BOSS Zhipin, architecture-focused frontend roles requiring 3+ years of experience average a salary of 32.5K RMB.
The skill transition path for a React developer is immediate: add AI API integration and streaming rendering to become an AI application frontend engineer.
Conclusions

The 165,000 layoff figure and the 23% hiring increase are not contradictory—they describe a single labor-market split between execution and judgment. The same force, AI, is simultaneously destroying and creating demand within one job title.

Frontend hiring criteria are shifting from tool proficiency to architectural judgment. Knowing React is now table stakes; the differentiator is whether you can evaluate AI-generated React code for correctness, performance, and edge cases.

The JD keyword shift from 'proficient use' to 'deep understanding of design principles' signals that employers now expect engineers to operate one level of abstraction higher than the code itself—exactly the layer AI cannot yet reach.

Salary stagnation for execution-type frontend roles and a 32.5K average for architecture-type roles suggest the market is already pricing the judgment premium. The gap will likely widen as AI coding tools improve.

The emergence of 'Full-Stack AI Engineer (Frontend Background)' as a distinct title indicates that prompt engineering and AI-output post-processing are becoming recognized frontend competencies, not backend or data-science side skills.

Concepts & terms
Execution-layer vs. Judgment-layer roles
Execution-layer roles take clear inputs (design mockups, API specs) and produce standardized outputs (pages, integrations)—work AI increasingly automates. Judgment-layer roles make architecture decisions, evaluate AI-generated output quality, and handle edge cases AI cannot resolve—work that demands human discernment and is growing in demand.
Streaming rendering
A frontend technique for displaying AI-generated text in real time as tokens arrive, rather than waiting for the full response. It requires handling Server-Sent Events or WebSockets, managing partial content state, and building fallback UIs for interrupted streams—a core skill in AI-integrated frontend roles.
AI output verification and quality control
The practice of reviewing AI-generated code or content for correctness, architectural soundness, performance issues, and edge cases. As AI writes more production code, the human role shifts from author to reviewer, making verification a primary frontend competency.
Source: juejin.cn ↗ Google Translate ↗ Backup ↗