1. At its peak in April 2020, approximately how many German workers were on the Kurzarbeit short-time work scheme?
Correct. Approximately 6 million German workers were on Kurzarbeit at the April 2020 peak — preventing mass layoffs while maintaining employer-employee relationships critical for any subsequent reskilling.
At its peak in April 2020, approximately 6 million German workers were on Kurzarbeit — the scheme that prevents temporary downturns from becoming permanent unemployment by sharing hours and government-compensating lost wages.
2. Bloomberg's BloombergGPT was a language model with approximately how many parameters?
Correct. BloombergGPT was a 50-billion parameter model trained specifically on financial data.
BloombergGPT had 50 billion parameters and was trained on financial domain data.
3. What was the key reason Walmart held store managers partially accountable for employee enrolment rates in Live Better U?
Correct. Embedding enrolment metrics in manager KPIs structurally addressed the voluntarism trap — turning what would otherwise be a low-participation optional programme into a managed outcome.
Walmart's design insight was that optional programmes fail structurally. Making managers accountable for enrolment rates converted voluntary participation into a managed outcome — the key fix for the voluntarism trap.
4. The concept of "client-specific relational knowledge" as a protective factor refers to which idea from Lesson 4?
Correct. Client-specific relational knowledge — the accumulated understanding of a particular organization's context, history, and personnel — is inherently slow to develop and cannot be compressed into a prompt, making it durable as AI capabilities grow.
Client-specific relational knowledge refers to the deep, context-specific understanding of a particular client or employer that accumulates slowly and cannot be replicated by AI without substantial access and onboarding — a form of knowledge that remains scarce as broadly-applicable skills become AI-accessible.
5. According to LinkedIn's 2023 data, what was the year-over-year growth rate in "AI Specialist" postings between 2021 and 2023?
Correct. 74% year-over-year growth.
LinkedIn reported 74% year-over-year growth in AI Specialist postings.
6. Scale AI's valuation in its 2021 funding round was approximately:
Correct. $7.3 billion in Scale AI's 2021 funding round.
Scale AI was valued at $7.3 billion in its 2021 funding round.
7. What does South Korea's 2022 amendment to the Employment Insurance Act require of large employers using AI automation above defined thresholds?
Correct. South Korea's amendment required higher Employment Insurance contributions from heavy automation users — placing a portion of displacement costs on the employing organisations causing displacement rather than socialising them through general taxation.
South Korea's 2022 amendment required employers above automation thresholds to pay higher Employment Insurance contributions — the policy principle being that those who cause displacement should internalise some of its costs.
8. Which of the following best describes a skills-based talent architecture?
Correct. Skills-based architecture replaces the job title as the unit of HR management with a skill profile — enabling far more flexible matching of people to evolving role requirements as AI changes what work requires.
Skills-based talent architecture organises HR around skill profiles rather than job titles — enabling fluid matching of people to changing role requirements, and surfacing internal supply that title-based systems would miss.
9. The Goldman Sachs 2023 analysis found that what percentage of current tasks are fully automatable using existing AI models?
Correct. Goldman Sachs found 25% of current tasks fully automatable, with two-thirds of all jobs exposed to some degree of AI automation.
Goldman Sachs's 2023 analysis found 25% of current tasks fully automatable with existing AI — and two-thirds of all jobs exposed to some degree of AI automation.
10. The 2023 TIME investigation documented what approximate hourly pay rate for Kenyan workers at Sama reviewing content for OpenAI?
Correct. $1.32–$2.00 per hour — raising significant labor conditions concerns.
TIME documented approximately $1.32–$2.00 per hour for this work in Nairobi.
11. What was the primary finding of the Brynjolfsson-Li-Raymond 2023 "Generative AI at Work" study regarding productivity effects?
Correct. The 14% average gain and novice-worker advantage are central findings. The AI taught novices strategies experienced workers already knew, compressing the skill gap in measurable ways.
The study found 14% average productivity gains, with novice workers benefiting most as the AI shared expert strategies they hadn't yet internalized — compressing the skill gap between novice and expert workers.
12. What is a "selection effect" in the context of reskilling program evaluation?
Correct. Selection effects make it difficult to know whether reskilling programs caused better outcomes, or whether workers who would have succeeded anyway are disproportionately participating.
A selection effect occurs when program participants aren't representative of all workers — more motivated, higher-skilled workers self-select in, inflating apparent program impact and making it hard to isolate the program's true effect.
13. What percentage of all jobs did the WEF Future of Jobs Report 2023 expect to change significantly within five years?
Correct. The WEF's 2023 report found 23% of all jobs expected to change significantly within five years across its surveyed companies.
The WEF's 2023 Future of Jobs Report found 23% of jobs expected to change significantly in the next five years across 803 surveyed companies.
14. According to LinkedIn's 2023 Workplace Learning Report, what percentage of job skills profiles globally were projected to change by 2030?
Correct. LinkedIn's 2023 report projected 65% change in skills profiles by 2030, having already measured 25% change since 2015.
LinkedIn projected 65% change by 2030 — the 25% figure was change already recorded since 2015, and 44% was the WEF's core skills change estimate for workers within 5 years.
15. The McKinsey Global Institute's 2023 report projected how many occupational transitions in the United States alone by the end of the decade?
Correct. McKinsey projected 12 million occupational transitions needed in the U.S. alone by the end of the decade.
McKinsey projected 12 million occupational transitions needed in the United States alone by the end of the decade — a figure driven by both AI-specific and broader digitization forces.
16. In the 70-20-10 learning model applied to AI-disrupted environments, what role does the 70% experiential component increasingly serve?
Correct. Researchers including Josh Bersin argue that AI-augmented work is itself a learning environment where real-time AI feedback accelerates experiential development.
In AI-disrupted environments, the 70% experiential component now includes working alongside AI tools as learning — AI assistants providing real-time feedback accelerate experiential skill development.
17. Which company published the "300 million jobs" estimate in March 2023, and what did the study project simultaneously?
Correct.
Goldman Sachs economists Briggs and Kodnani published the 300M figure alongside a 7% global GDP growth projection — context that rarely accompanied the headline.
18. GPT-4 scored in approximately what percentile on the Uniform Bar Examination in March 2023?
Correct. GPT-4 scored in the 90th percentile on the UBE in March 2023.
GPT-4 scored in the 90th percentile — passing comfortably but triggering hybrid role creation rather than lawyer displacement.
19. What specific quality improvement did BCG observe when consultants used GPT-4 assistance on business tasks in its 2022 study?
Correct. BCG found AI-assisted consultants produced work judged 40% higher quality, completed tasks 25% faster and finished 12% more tasks — alongside the critical caveat about over-reliance outside AI's competency.
BCG measured a 40% quality improvement for AI-assisted work — alongside a 25% speed increase and 12% more tasks completed — but with the important caveat that over-reliance degraded performance outside AI's competency.
20. The ATM case shows that automation of the cash-dispensing task led to what labor market outcome for bank tellers by 2010?
Correct.
ATMs reduced branch operating costs, enabling more branches, which required more tellers for sales and relationship work — net teller employment rose.