DW Shift — how AI is changing the working world
▶ Watch on YouTube · DW Shift · 12:31
How is artificial intelligence transforming hiring, entry-level roles, and the future of work? Get ready for a world where AI doesn’t just help you do your job, it may even decide whether you get one.
TL;DR
A 12-minute report on DW Shift, Deutsche Welle’s technology channel, published 17 October 2026. A presenter-led piece with short interviews, mostly in Germany and London. Five threads:
- Expectations of job cuts have fallen. In an EY survey of CEOs, 46% expected AI to lead to significant job cuts in early 2025; by May 2026, 20%.
- A digital twin for every employee. Technology consultant Richard Skellett has fed four years of his data (photos, bank accounts, e-mails, every meeting) into “Digital Richard,” which colleagues and clients can talk to and which “cannot go off and do anything without human Richard.” Every employee of his firm is offered one. The open question: who owns a twin when the employee leaves? Skellett’s answer is the employee, with the company paying a fee to keep using it.
- Entry-level work is harder to find. A Dresden computer-science graduate with a master’s degree and work experience sent more than 300 applications before finding a job. “There are fewer entry-level jobs and employers are now looking for different skills.”
- AI is on both sides of hiring. Candidates use AI for CVs and cover letters; employers use it to write job ads, screen applications and run first interviews through avatars that also check identity by face scan. A London recruiter: employers now “ask to put more human back into the process.”
- Creative work is already losing. A voice actor says AI is costing him and his industry “a lot of work” in advertising and games; a Berlin stylist lost 70% of her work when her client switched to AI retouching.
Key claims
The digital twin
Digital Richard describes its own role: “My job is to amplify Richard’s original intelligence… I can challenge his thinking, but I can not take his place.” Skellett sets its guardrails and decides “what I learn, what I’m allowed to do and which decisions need his approval.” He reports “an enormous increase in productivity” from twins that are “fast, are never sick and work around the clock,” and names the unresolved questions: who owns the twin, what its output may be used for, and whether a former employer may keep using it after the employee joins a competitor. His son uses Digital Richard for maths homework; Skellett calls it “not artificial intelligence. It’s my original intelligence.”
Entry level, universities, and hiring
Routine programming is increasingly done by generative AI, so a newly trained engineer who specialised in traditional programming “would find it more difficult to keep up” and must learn what AI can do now. A university voice in the piece: “Actually, everything needs to change,” with AI explaining concepts “again and again” as a chance to learn more personally. On hiring, a clear CV matters more because systems read applications automatically, but AI “does not understand a person in the same way a human recruiter does,” and candidates interviewed by an avatar say they miss “an opportunity really to present themselves.”
Creative work
China has a full AI film industry with its own festivals, and an AI “actor,” Tilly Norwood, asks to be taken seriously. Voice actor Christoph Walter expects to lose his job and is building a second career in marketing; an AI voice still needs a human model for how a person “pauses… breathes… uses a certain accent.” An unnamed expert cautions that people “jump ahead to the labor market disruption is happening right now.” The closing advice: “It’s not humans OR AI. It’s humans working WITH AI,” and build “judgement, creativity, and connection.”
How it connects
- ai-employment-effects — the falling share of CEOs expecting cuts, the entry-level squeeze, AI in recruiting, and the first creative occupations losing work.
- Sundararajan (WEF Radio Davos) — the entry-level slowdown with its cause left open.
- Fowler (Bloomberg Tech) — the same week, the case for universities from a fund that backs students.
What was actually ingested
The full report (12:31, length_seconds: 751) from the human-curated caption track, fetched with yt-dlp after the transcript panel failed. Most speakers are unnamed in the captions; quotes are attributed only where the captions name the speaker. Not tagged with dynamic_capabilities: it is reportage on the labour market rather than on how organisations adopt AI. No stills: not probed.
Linked entities and concepts
- DW News — Deutsche Welle’s entity; DW Shift is added there as an alias. Updated in this ingest.
- ai-employment-effects — updated in this ingest.
Dangling (single-source mention, deferred): Richard Skellett. EY was promoted to an entity page on 2026-10-25.
Scope and reliability
A short television report built from individual cases. The EY figures are the only survey data, cited without method. The graduate, the voice actor and the stylist are single cases, chosen to illustrate; the stylist’s 70% is her own account. The digital-twin productivity claim is the twin owner’s. Its value is the range of cases in one place (twins, entry level, AI on both sides of hiring, creative work) and the ownership question for digital twins, which no other source in the wiki raises.