Goodwill CEO Preston Taps AI Tools to Cut Recidivism, Fill 2.6M Jobs

Fireside Chat with Steve Preston, Goodwill with Thirty Minute Mentors and Goodwill Industries — Ai4

Ai4

Targeted AI assessments and personalized workforce training deliver asymmetric economic returns over incarceration by steering overlooked populations directly into critical labor shortages.

AI-assisted summary of Fireside Chat with Steve Preston, Goodwill with Thirty Minute Mentors and Goodwill Industries — original source. Check the original for full context.

Source published: 2026-08-22

Detailed notes

## Background Published by Ai4 on YouTube on August 22, 2026, this fireside chat features Adam Mendler interviewing Goodwill Industries International President and CEO Steve Preston. The session explores the significant AI training and literacy gaps in the modern workforce, Goodwill's operational adoption of artificial intelligence, and how both companies and individuals can navigate transformative career pathways.

## TL;DR Goodwill CEO Steve Preston outlines how AI and data-driven interventions can bridge the divide between more than 40 million underemployed Americans and millions of open jobs in logistics, healthcare, and skilled trades. He explains how Goodwill deploys AI to streamline diagnostic intake, barrier assessments, and personalized learning, freeing frontline career coaches to provide high-touch human mentorship and wraparound social support. Highlighting evidence-based successes—such as reducing recidivism rates to the mid-single digits—Preston urges corporate HR leaders, technologists, and investors to partner with community workforce developers, framing the upskilling of overlooked talent as an extraordinary financial and societal return on investment.

## Atmosphere & style The discussion carries a measured, pragmatic, and purpose-driven tone, seamlessly blending high-level economic analysis with grounded humanitarian empathy. Steve Preston communicates with the sharp financial discipline of an experienced public servant and former investment banker while maintaining a steadfast focus on vulnerable job seekers. The exchange feels both authoritative and deeply inspiring, reframing AI from an intimidating source of labor disruption into an essential diagnostic and developmental catalyst for human mobility.

## Key takeaways - [00:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=4s) Discussions surrounding AI have evolved beyond technical capabilities into a human-centered workforce transition focused on creating sustainable economic outcomes. - [02:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=124s) A massive labor market mismatch persists in the United States, with over 40 million underemployed or low-wage workers alongside millions of unfilled roles in healthcare, logistics, and skilled trades. - [04:18](https://www.youtube.com/watch?v=dn75IsD4gD8&t=258s) AI diagnostics can instantly evaluate work readiness and life barriers, enabling tailored learning pathways and automated social service referrals. - [04:18](https://www.youtube.com/watch?v=dn75IsD4gD8&t=258s) Rather than displacing human counselors, AI-enabled administrative automation frees career coaches to focus on high-touch behavioral mentorship and employer readiness. - [08:07](https://www.youtube.com/watch?v=dn75IsD4gD8&t=487s) The workforce development tech ecosystem remains fragmented, requiring scale and connective data infrastructure to determine which interventions work best. - [08:07](https://www.youtube.com/watch?v=dn75IsD4gD8&t=487s) Data-informed wraparound interventions have proven capable of driving justice-involved recidivism down into the mid-single digits compared to the 30% to 40% national average. - [12:05](https://www.youtube.com/watch?v=dn75IsD4gD8&t=725s) A symbiotic economic loop exists where equipping marginalized individuals with career pathways directly addresses acute corporate labor shortages. - [13:02](https://www.youtube.com/watch?v=dn75IsD4gD8&t=782s) Talent and human potential are evenly distributed across society, but access to training, technology, and career opportunities is not. - [13:02](https://www.youtube.com/watch?v=dn75IsD4gD8&t=782s) Technology vendors and business leaders should partner directly with frontline workforce nonprofits to pilot AI solutions on practical, ground-level challenges. - [15:40](https://www.youtube.com/watch?v=dn75IsD4gD8&t=940s) The return on investment for upskilling and rehabilitation dramatically outweighs the compounding public liabilities of incarceration and generational poverty.

## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=4s) Introduction and Expanding AI Beyond Pure Technology

Adam Mendler introduces Steve Preston, highlighting his diverse leadership career spanning both the private and public sectors, including his service as head of the Small Business Administration (SBA) and as a presidential cabinet secretary (leading the Department of Housing and Urban Development), prior to becoming the President and CEO of Goodwill Industries International—the largest workforce development organization in the United States.

When asked about his unique perspective on AI and how Goodwill utilizes it, Preston observes that discussions around AI have matured from early debates about pure technical capabilities to a much deeper conversation about "what it means for people, what it means for humanity, what it means for how we develop ourselves." Managing an organization of 140,000 employees alongside a vast retail network and workforce development infrastructure, Preston explains that Goodwill's core mission is understanding how AI intersects with employers, workers, and society at large. While upskilling people with tactical AI proficiencies is vital, he stresses that the ultimate priority is navigating a massive, multifaceted workforce transition to secure better economic outcomes for communities and families.

[02:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=124s) Addressing the Macro Workforce Imbalance

Preston frames the AI transition within a broader context of shifting demographics, an aging population, and deep systemic labor market distress. Across the United States, over 40 million people are unemployed, underemployed, or stuck in very low-wage jobs. Simultaneously, severe labor shortages persist in critical sectors, with nearly 2 million open roles across healthcare and logistics, alongside well over 600,000 vacancies in skilled trades and building maintenance.

```mermaid flowchart TD accTitle: AI-Driven Workforce Transition Framework subgraph S1["Workforce Challenges"] A["40M+ Unemployed, Underemployed, & Low-Wage Workers"] B["Demographic Shifts & Aging Population"] end

subgraph S2["AI-Enabled Interventions"] C["Aggregated Diagnostic & Barrier Assessments"] D["Adaptive Learning & Personalized Training"] E["Integrated Wraparound Support (Housing/Childcare)"] end

subgraph S3["Target Outcomes"] F["Filling 2M+ Openings in Healthcare & Logistics"] G["Filling 600K+ Openings in Skilled Trades & Maintenance"] H["Sustainable Wage Growth & Reduced Recidivism"] end

S1 --> S2 S2 --> S3 ```

Preston argues that these vacancies represent high-quality careers capable of changing life trajectories if underrepresented populations can receive the right training. AI can accelerate this transition by equipping individuals with support tools and capturing comprehensive data to measure service efficacy. Beyond macro initiatives, Goodwill has already provided foundational AI tools and skills training directly to roughly 25,000 individuals.

- **40+ million**: Number of Americans currently unemployed, underemployed, or earning low wages. - **2+ million**: Open jobs in healthcare and logistics needing qualified workers. - **600,000+**: Unfilled openings in skilled trades and facility maintenance. - **25,000**: Individuals directly trained by Goodwill in AI tools and core competencies.

[04:18](https://www.youtube.com/watch?v=dn75IsD4gD8&t=258s) Designing the AI-Augmented Career Coaching Journey

To illustrate AI's practical potential, Preston walks through the experience of a typical job seeker entering a Goodwill center. Approximately 85% to 86% of the individuals Goodwill serves have a high school diploma or less, and many navigate severe hurdles, such as long-term unemployment or prior justice involvement. In a standard workflow, a career coach administers multiple assessments evaluating life stability, work readiness, and technical skills, matching those findings against real-time local labor market data to construct a customized development plan.

Under an AI-augmented model, the aggregation of diagnostic assessments, barrier identification, and customized pathway generation happens almost instantly. If an individual struggles with housing instability, childcare gaps, or work-readiness deficits, AI tools can streamline referrals across third-party social service networks and monitor participation. Furthermore, online coursework can incorporate adaptive, one-on-one personalized learning. Rather than replacing human counselors, these automated workflows liberate career coaches from administrative burdens, empowering them to deliver high-touch mentorship, behavioral coaching, and employer-backed job readiness preparation.

[08:07](https://www.youtube.com/watch?v=dn75IsD4gD8&t=487s) Scaling Innovation, Ecosystem Guts, and Predictive Analytics

Addressing how close the field is to realizing this comprehensive vision, Preston notes that while small innovators and major entities—such as Coursera's investments in adaptive learning, alongside philanthropic backing from the Gates Foundation, the Ballmer Group, and Stand Together—are engineering point solutions, the broader ecosystem remains fragmented. Crucial connective tissue between specialized tools is still missing, which is why Goodwill leverages its scale to test and co-develop emerging technologies.

```mermaid flowchart LR accTitle: Goodwill Data Aggregation and Feedback Engine subgraph Input["Population Scale"] A["2M People Touched / Year"] B["200K+ Intensive Cases"] end subgraph Engine["AI Aggregation & Analytics"] C["Cross-Platform Data Integration"] D["Predictive Intervention Modeling"] end subgraph Stakeholders["Ecosystem Beneficiaries"] E["Frontline Coaches"] F["Policymakers & Researchers"] G["Employers & Candidates"] end

Input --> Engine Engine --> Stakeholders ```

Preston emphasizes that Goodwill serves over 2 million people annually, delivering intensive, individualized career support to more than 200,000 individuals. This footprint positions the organization to build data aggregation systems that identify precisely which interventions produce positive outcomes for specific life challenges. Drawing an analogy to clinical medical diagnostics, he notes that data-backed decision support can drastically reshape social outcomes. For instance, while national recidivism rates for justice-impacted individuals typically hover between 30% and 40% within the first two years post-release, Goodwill's dedicated, data-informed justice-involved programs have driven recidivism down into the mid-single digits.

- **2 million**: Total individuals assisted by Goodwill programs each year. - **200,000+**: Individuals receiving dedicated, intensive one-on-one career case management annually. - **30% to 40%**: Typical national baseline recidivism rate within 1–2 years of release from incarceration. - **Mid-single digits**: Recidivism rate achieved within Goodwill's intensively tracked, evidence-based justice programs.

[12:05](https://www.youtube.com/watch?v=dn75IsD4gD8&t=725s) Unlocking Pathways Out of Poverty and Fueling Economic Growth

Reflecting on targeted pilot programs, Steve Preston highlights how combining holistic wraparound services with structured support leads to "astoundingly low recidivism rates." While these successes currently exist within specific program microcosms, scaling that level of insight across all workforce development initiatives would create a fundamentally different pathway out of poverty and hardship. Rather than viewing social support in isolation, Preston stresses that this pathway directly meets the desperate hiring demands of businesses: it fuels the broader economy by supplying critical talent to employers facing acute labor shortages.

This alignment between social support and market demand represents "the most symbiotic thing you can imagine" in contemporary society. To realize this potential at scale, organizations must bridge the gap between emerging artificial intelligence tools and deeper organizational understanding powered by cleaner, comprehensive data. Preston emphasizes that building these analytical bridges creates a massive unlock for workforce providers, job seekers, and the business community alike.

- **The Symbiotic Value Loop**: Supporting vulnerable job seekers concurrently lifts individuals out of poverty and resolves critical talent deficits for growing enterprises. - **Data and AI as Catalysts**: Microcosm-level interventions show dramatic drops in recidivism, demonstrating that scaling these outcomes requires integrating predictive data and modern AI tools across social services.

[13:02](https://www.youtube.com/watch?v=dn75IsD4gD8&t=782s) Actionable Advice for HR, Tech Partners, and Business Leaders

Addressing an audience spanning startup entrepreneurs, corporate executives, independent contributors, and machine learning specialists, Adam Mendler asks how diverse stakeholders can get practically involved. Preston directs his first recommendation to HR leaders and corporate recruiters, urging them to recognize the massive, overlooked talent pool waiting across the nation. Emphasizing that "talent and potential is evenly distributed; opportunity is not," he notes that systemic lack of access and life hurdles prevent capable workers from entering traditional pipelines. Companies must rethink how they source, train, develop, and nurture candidates, intentionally partnering with non-traditional workforce developers like Goodwill to tap into unconventional pipelines.

For technology professionals, IT service providers, and venture investors, Preston advocates actively steering technical resources toward nonprofit and community challenges. Rather than treating social impact as a separate philanthropic silo, organizations should consult with their IT teams and external advisers to explore where AI tools can streamline operations and improve client outcomes. By collaborating directly with local workforce organizations to pilot innovative tools in real-world environments, developers can discover what works on the ground and accelerate the national dialogue on talent development.

```mermaid flowchart TD accTitle: Goodwill Collaborative Talent Model subgraph Partners["Cross-Sector Stakeholders"] HR["HR & Corporate Recruiters"] TECH["Tech Vendors & IT Partners"] INV["Investors & Philanthropists"] end subgraph Pipeline["Nontraditional Workforce Pipeline"] GII["Goodwill & Community Nonprofits"] TRAIN["Targeted Training & AI Tooling"] WRAP["Wraparound Support Services"] end subgraph Market["Economic & Social Impact"] DIV["Equitable Opportunity Access"] LABOR["Skilled Talent for Employers"] end

Partners --> GII GII --> TRAIN TRAIN --> WRAP WRAP --> Market ```

- **Rethinking Candidate Pipelines**: HR departments are encouraged to look beyond standard resume profiles and collaborate with community developers who cultivate resilient talent. - **Engaging Technical Capabilities**: IT departments and technology consultants can partner with local nonprofits to deploy and stress-test operational AI solutions in frontline environments.

[15:40](https://www.youtube.com/watch?v=dn75IsD4gD8&t=940s) Changing the Narrative: The Irresistible ROI of Human Potential

Turning to the broader cultural perception of disadvantaged communities, Preston urges a fundamental shift in the narrative around what marginalized individuals can achieve. It is easy for passersby to look at struggling urban neighborhoods or individuals re-entering society after prison and resign themselves to skepticism. Yet daily operations across Goodwill disprove that pessimism: individuals starting from intensely difficult situations regularly transform their trajectories within months, overcoming formidable barriers to secure positions at "some of the most storied organizations in our country." Realizing this widespread change requires three core commitments: believing transformation is possible, acknowledging that the investment is worthwhile, and actively dismantling systemic obstacles that stall progress.

To substantiate the economic case, Preston—drawing on his past career as an investment banker—frames social rehabilitation in terms of financial return on investment. Incarcerating an individual costs upwards of $40,000 per year, not accounting for the secondary economic devastation of an absent parent, dependent family assistance, and the multi-generational churn of the criminal justice system. Comparing those compounding liabilities against the modest cost of technological and skills-based intervention reveals an extraordinary dividend: individuals gain long-term self-sufficiency, children secure stable domestic foundations, and employers gain loyal workers. Preston notes this constitutes an ROI that "any investment banker or private equity person in New York or in the valley would take any time any day of the week."

```mermaid flowchart TD accTitle: The Societal and Economic ROI of Workforce Reentry subgraph Inaction["Cost of Systemic Inaction"] C1["Incarceration Costs: $40,000+ per Year"] --> C2["Parental Absence & Family Destabilization"] C2 --> C3["Multi-Generational Churn & Lost Economic Output"] end subgraph Reentry["Targeted Intervention & AI Enablement"] I1["Holistic Support & Technological Upskilling"] --> I2["Placement in Resilient, Career-Track Roles"] end subgraph Dividend["Financial & Societal Return"] R1["Self-Sufficient Taxpaying Households"] R2["Stable Upward Trajectory for Children"] R3["Direct Labor Supply for Businesses"] end

Reentry --> Dividend ```

- **Direct Incarceration Costs**: Housing an incarcerated individual exceeds $40,000 annually in public funds, excluding ancillary downstream expenditures. - **Multi-Generational Impact**: Removing a parent creates ripple effects that disrupt child welfare and household financial stability for decades. - **The Financial and Human Bottom Line**: The capital required to upskill a worker delivers asymmetric economic returns compared to the cost of systemic neglect, positioning AI tools as a critical mechanism to unlock human capability nationwide.

## Notable quotes > [00:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=4s) what it means for people, what it means for humanity, what it means for how we develop ourselves. — Steve Preston

> [12:05](https://www.youtube.com/watch?v=dn75IsD4gD8&t=725s) astoundingly low recidivism rates — Steve Preston

> [12:05](https://www.youtube.com/watch?v=dn75IsD4gD8&t=725s) the most symbiotic thing you can imagine — Steve Preston

> [13:02](https://www.youtube.com/watch?v=dn75IsD4gD8&t=782s) talent and potential is evenly distributed; opportunity is not — Steve Preston

> [15:40](https://www.youtube.com/watch?v=dn75IsD4gD8&t=940s) some of the most storied organizations in our country — Steve Preston

> [15:40](https://www.youtube.com/watch?v=dn75IsD4gD8&t=940s) any investment banker or private equity person in New York or in the valley would take any time any day of the week. — Steve Preston

## Mentioned Steve Preston, Adam Mendler, Goodwill Industries International, Thirty Minute Mentors, Ai4, Small Business Administration (SBA), Department of Housing and Urban Development, Coursera, Gates Foundation, Ballmer Group, Stand Together

## Video description Adam Mendler - Creator & Host, Thirty Minute Mentors Steve Preston - CEO, Goodwill Industries Internationa

As AI transforms industries, organizations, and jobs, what can individuals do to stay relevant, and what role should companies play in ensuring employees are AI literate? A recent report from Goodwill Industries International (GII) indicates that only 22 percent of Americans use AI for career training and upskilling, while 35 percent report not using AI professionally at all. Forty-nine percent of respondents said they cannot keep up with the pace at which AI is moving, and only 40% said their employers provide adequate AI training. As GII President and CEO, Steve Preston leads both the largest nonprofit workforce development provider in the country and the largest network of thrift stores. In this fireside chat, Steve will talk about the current AI gaps in the workforce, the pathways we need to create for individuals to bridge to the AI future, how Goodwill is using AI to improve operations, what companies should do to ensure a competitive workforce in the AI world, and how individuals can leverage this pivotal period to advance their own careers.

[Source](https://www.youtube.com/watch?v=dn75IsD4gD8)

中文简报

Goodwill 首席执行官借助 AI 工具降低再犯率并填补 260 万岗位空缺针对性的 AI 评估和个性化劳动力培训将容易被忽视的人群直接输送至关键劳动力短缺领域，相比监禁带来了不对称的经济回报。 Goodwill 首席执行官 Steve Preston 阐述了人工智能和数据驱动的干预措施如何弥合 4000 多万就业不足的美国人与物流、医疗和技术工种数百万空缺职位之间的差距。他解释了 Goodwill 如何部署 AI 来简化诊断接案、障碍评估和个性化学习流程，从而让一线职业导师得以腾出精力，提供高互动性的人文指导与全方位的社会支持。Preston 强调了以数据为支撑的成功案例（例如将再犯率降低至个位数中段），并呼吁企业人力资源主管、技术人员和投资者与社区劳动力开发机构合作，将被忽视人群的技能提升视为一项能带来巨大财务与社会回报的投资。围绕 AI 的讨论已超越纯技术能力范畴，演变为以人为本的劳动力转型，重点在于创造可持续的经济成果。 (0:04) 美国劳动力市场依然存在严重的供需错配，在超过 4000 万就业不足或低薪工人的同时，医疗、物流和技术工种仍有数百万职位空缺。 (2:04) AI 诊断工具可即时评估就业准备情况与生活障碍，从而支持定制化学习路径和自动化社会服务转介。 (4:18) AI 带来的行政流程自动化并未取代人类咨询师，反而让职业导师能够专注于高互动性的行为指导和求职准备培训。 (4:18) 劳动力开发技术生态系统仍处于割裂状态，需要扩大规模并建立连通性的数据基础设施，以验证哪些干预措施最为有效。 (8:07)

Reading summary

Steve Preston, President and CEO of Goodwill Industries International, argues that integrating AI tools and predictive analytics into workforce development can bridge acute national labor shortages by transitioning overlooked, vulnerable workers into sustainable, high-demand careers. Bridging Macro Labor Shortages with Underrepresented Workers (2:04) Preston frames the workforce challenge around an aging population and structural labor imbalances, pointing out that over 40 million Americans remain unemployed, underemployed, or stuck in low-wage jobs. Simultaneously, critical industries face severe talent shortages, including over 2 million open roles across healthcare and logistics, as well as more than 600,000 vacancies in skilled trades and facility maintenance. Contending that these unfilled positions offer life-changing career pathways if workers receive appropriate support, Preston notes that Goodwill has already trained roughly 25,000 individuals directly in foundational AI tools and core digital competencies to help accelerate their entry into these resilient sectors. Automating Administration to Enhance Human Coaching (4:18) Approximately 85% to 86% of individuals entering Goodwill centers have a high school diploma or less, with many navigating compounding hurdles such as long-term unemployment, past justice involvement, or housing instability. In traditional settings, career coaches must spend substantial time administering multiple diagnostic assessments, evaluating work readiness, and manually identifying supportive community services. Under an AI-augmented model, assessment aggregation, barrier identification, and customized pathway generation occur almost instantly. Automated tools streamline external referrals for childcare or housing and deliver personalized, adaptive learning modules. Rather than replacing human counselors, Preston emphasizes that these automated workflows liberate staff from routine administrative duties, enabling them to provide high-touch mentorship, behavioral coaching, and job-readiness preparation. Scaling Predictive Interventions and Lowering Recidivism (8:07) While organizations like Coursera and philanthropies such as the Gates Foundation, Ballmer Group, and Stand Together are funding point innovations, Preston notes that the overall workforce technology landscape remains fragmented without connective infrastructure. Goodwill leverages its operational footprint—assisting over 2 million individuals annually, including more than 200,000 through intensive case management—to build data aggregation engines that analyze which specific interventions work best for particular barriers. Drawing a comparison to clinical medical diagnostics, Preston asserts that predictive data models can dramatically shift social outcomes. While national recidivism rates for justice-involved populations typically range between 30% and 40% within the first two years following release, Preston states that Goodwill's data-informed support programs have reduced recidivism down into the mid-single digits. Employer Partnerships and the Economic Case for Reentry (13:02) Drawing on his background as an investment banker, Preston highlights the substantial financial return on investment from workforce rehabilitation compared to the compounding costs of inaction. Directly incarcerating an individual costs more than $40,000 annually, alongside the downstream societal and economic damage of parental absence and generational churn. Investing in training and wraparound support instead yields self-sufficient taxpayers and fills critical talent gaps for businesses. To scale these results, Preston advises HR leaders and recruiters to look past conventional resumes and intentionally partner with community workforce developers. He also urges technology specialists and investors to deploy operational AI and analytical capabilities directly within nonprofit environments to pilot real-world solutions. Preston concludes that creating a scalable path out of poverty while resolving employer talent shortages requires breaking down sector silos, connecting fragmented software point solutions with clean predictive data, and actively testing AI-driven interventions in frontline community environments.

精炼阅读摘要

Goodwill Industries 总裁兼首席执行官 Steve Preston 指出，人工智能在劳动力发展中的核心价值并非单纯的技术替代，而是通过自动化行政负担、整合综合数据与支持个性化培训，帮助数千万处于困境中的求职者填补关键行业的劳动力缺口，从而实现显著的社会与经济回报。宏观用工缺口与弱势求职者的结构性矛盾 (2:04) 全美目前有超过 4000 万人处于失业、就业不足或从事极低薪工作的状态。与此同时，医疗保健和物流领域的空缺岗位接近 200 万个，技术工种和设施维护岗位的空缺也超过 60 万个。Preston 认为，随着人口结构变化和人口老龄化加剧，劳动力市场面临深层次的系统性困境，而这些空缺岗位本身能够成为改变弱势群体生活轨迹的高质量职业通道。在 Goodwill 服务的对象中，约 85% 至 86% 的人最高学历仅为高中或以下，许多人还面临长期失业或有犯罪记录等严峻障碍。虽然该机构已直接向约 25,000 人提供了基础人工智能工具和技能培训，但宏观层面的根本挑战在于如何通过精准的技能提升与过渡支持，将未被充分利用的人力资源对接至面临严重用工短缺的关键行业。 AI 赋能的职业指导与个性化支持网络 (4:18) 在传统的职业指导流程中，求职者进入中心后需要接受多项生活稳定性、工作准备度和技术能力的评估，顾问需耗费大量时间进行行政协调。而在 AI 增强模型下，系统能够近乎即时地完成诊断评估、识别求职障碍并生成定制化路径规划。若求职者面临住房不稳定或育儿资源匮乏等困难，AI 工具可快速跨第三方社会服务网络发起转介并跟踪进展。自动化工作流程的引入旨在将职业顾问从烦琐的文书工作中解放出来，从而有更多精力提供高互动性的导师辅导、行为指导以及雇主认可的就业准备支持。此外，在线课程能够结合自适应的一对一个性化学习，在不替代人类咨询师的前提下，显著优化求职者的技能获取效率与干预精准度。利用数据飞轮与预测分析降低再犯率 (8:07) Goodwill 每年服务超过 200 万人，并为超过 20 万人提供深度的一对一职业个案管理。Preston 表示，这种庞大的服务规模使其能够构建跨平台的数据集成与预测分析系统，精确追踪哪些干预措施能针对特定的生活障碍产生积极结果，其逻辑类似于临床医疗诊断中的决策支持。这种循证干预的成效在司法涉入群体的帮扶中尤为突出。全美刑满释放人员出狱后 1 至 2 年内的基准再犯率通常在 30% 至 40% 之间，而在 Goodwill 实施精细数据追踪和全方位支持的专属项目中，再犯率被成功压低至中个位数（mid-single digits）。不过 Preston 也指出，尽管 Coursera、盖茨基金会、Ballmer Group 和 Stand Together 等机构在自适应学习或资金方面进行了单点探索，整个生态系统目前仍然处于碎片化状态，工具之间的关键连接机制仍有待健全。重塑招聘渠道与社会投资的投资回报率针对企业人力资源部门和技术开发者，Preston 呼吁破除对传统履历的依赖，主动与非营利性劳动力开发机构合作以开拓被忽视的人才库，同时建议科技团队直接与一线社区组织联合测试实际运营场景中的 AI 工具。他强调潜在的人才与潜力是均匀分布的，但获得机遇的途径却不均等，支持弱势求职者既能帮其摆脱贫困，又能直接解决企业的人才荒，形成高度共生的价值循环。从经济回报的角度考量，Preston 引用其过往的投资银行经验指出，关押一名服刑人员每年需消耗超过 40,000 美元的公共资金，这还不包括父母缺位对家庭及下一代造成的长期经济代价。相比之下，为个人提供技术与技能培训的投入相对有限，却能换取个人长期自立、家庭稳定并为雇主输送忠诚员工。这种非对称的经济与社会回报，其吸引力足以匹敌任何私募股权或风险投资项目。 Steve Preston 总结认为，改变边缘群体的经济轨迹需要全社会转变认知，打破系统性障碍，并承认对人力潜能的投资具有极高价值。然而，要将局部试点取得的惊人低再犯率和就业成效推广至全美范围，仍取决于整个劳动力发展生态能否克服工具碎片化瓶颈，建立起基于清洗后全面数据与现代 AI 工具的底层互联体系。

Detailed notes — zh

## Background 本内容由 YouTube 频道 Ai4 于 2026 年 8 月 22 日发布，记录了《Thirty Minute Mentors》创作者兼主持人 Adam Mendler 与 Goodwill Industries International 总裁兼首席执行官 Steve Preston 的炉边对谈。视频围绕劳动力转型展开，探讨了当前劳动力市场中的 AI 技能鸿沟、企业赋能员工的责任，以及 Goodwill 如何借助 AI 优化服务并帮助个体跨越职业发展障碍。 ## TL;DR 在本次对谈中，Steve Preston 深入剖析了全美超 4000 万低薪或失业人群与数百万核心岗位用工荒并存的结构性矛盾，指出 AI 在赋能弱势群体与优化劳动力服务体系中的双重价值。通过介绍 Goodwill 依托数字化路线图、自适应学习与精准干预将刑释人员重犯率大幅降至中个位数的实践，他展示了数据驱动干预的巨大潜力。最后，他从投资回报率（ROI）的视角呼吁企业管理者、HR 与技术界打破认知偏见，将 AI 视为解决劳动力供需错配与斩断代际贫困的关键杠杆，实现社会效益与经济回报的深度共生。 ## Atmosphere & style 整场对话兼具务实的战略高度与深切的人文关怀，呈现出沉稳、严谨且极具启发性的专业氛围。Steve Preston 凭借横跨政商与非营利领域的丰富履历，将宏观的商业财务思维（如精准计算的社会 ROI）与基层一线对弱势个体的共情完美融合，表达既客观笃定又充满感染力。主持人与嘉宾之间的交流流畅自然、直击要害，让听众在理性的数据论证与生动的案例剖析中，深切感受到技术向善撬动社会系统性变革的巨大希望。 ## Key takeaways - [00:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=4s) AI 行业的讨论焦点正从单纯的技术能力演变为对人类福祉、社会转型与人才发展路径的深远影响。 - [02:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=124s) 全美面临超 4000 万失业或低薪人群与医疗、物流、技工等领域逾 260 万岗位空缺并存的结构性供需矛盾。 - [02:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=124s) AI 在劳动力转型中具备双重赋能路径：一方面为个体提供自适应学习与辅助工具，另一方面通过聚合全流程数据大幅提升服务体系效率。 - [04:18](https://www.youtube.com/watch?v=dn75IsD4gD8&t=258s) AI 能整合评估数据并自动联动住房、心理健康与托儿等第三方支持网络，将职业导师从繁杂行政中解放出来以专注于深度共情辅导。 - [08:07](https://www.youtube.com/watch?v=dn75IsD4gD8&t=487s) 劳动力发展领域当前处于工具创新的“点状分布”阶段，亟需构建底层连接纽带以打通各单点解决方案。 - [08:07](https://www.youtube.com/watch?v=dn75IsD4gD8&t=487s) AI 最具突破性的潜力在于大规模跨周期数据聚合，可实现类似“现代医学精准诊断”的劳动力干预体系。 - [08:07](https://www.youtube.com/watch?v=dn75IsD4gD8&t=487s) 依托精准研究与深度个案介入，Goodwill 成功将刑释人员的重犯率从全美平均的 30%-40% 降至中个位数（约 4%-6%）。 - [12:05](https://www.youtube.com/watch?v=dn75IsD4gD8&t=725s) 通过 AI 与数据赋能弱势群体脱贫并填补企业紧缺用工，构成了当今社会最完美的“共生关系”。 - [13:02](https://www.youtube.com/watch?v=dn75IsD4gD8&t=782s) 企业人才管理者应打破偏见，认识到“潜能是均匀分布的，但机会并非如此”，主动拓宽非传统人才的选拔与培养渠道。 - [15:41](https://www.youtube.com/watch?v=dn75IsD4gD8&t=941s) 从综合投资回报率（ROI）考量，主动投资弱势群体技能培训与生活重塑的经济及社会收益，远优于放任其陷入司法系统或福利救济的巨额社会成本。 ## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=4s) 嘉宾背景与 AI 对劳动力转型的宏观意义主持人 Adam Mendler 开场介绍了 Goodwill Industries International 总裁兼首席执行官 Steve Preston。Steve Preston 拥有横跨公共部门与私营领域的深厚履历，不仅领导过多间私营及上市公司，还曾担任美国内阁部长（住房和城市发展部部长）以及美国中小型企业管理局（SBA）局长。在谈及当下对 AI 的看法时，Steve Preston 指出，AI 是一个层次极其丰富的议题。行业讨论在不久之前还停留在技术本身“能做什么”，而如今核心议题已经演变为 AI 对人类个体、社会福祉以及人才培养路径意味着什么。Goodwill 拥有 14 万名员工，旗下兼具庞大的二手零售连锁网络与全美最大的非营利劳动力发展体系。在如此庞大的组织体量下，Goodwill 关注的核心不仅是为劳动力传授 AI 工具与基础技能，更是从更宏观的视角审视整个社会正在经历的剧烈劳动力结构转型——思考如何利用 AI 作为杠杆，为劳动者、用人单位和整体经济带来真正向好的改变。 --- ### [02:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=124s) 劳动力市场供需错配与 AI 介入的双重抓手当前劳动力市场正面临深层次的结构性矛盾：人口老龄化加剧、人口结构持续演变，全美有超过 4000 万人处于失业、不充分就业或仅能赚取极低工资的困境中；与此同时，就业市场端却面临严重的用工荒——医疗保健与物流行业存在近 200 万个职位空缺，技术工种（Trades）及建筑维护等领域也有超过 60 万个用人缺口。这些岗位大多是薪资可观的优质工作，如果能对处于困境中的人群进行针对性技能培训，不仅能彻底改变个人及其家庭的命运轨迹，还能为企业输送急需人才并促进经济发展。 ```mermaid flowchart TD accTitle: AI 赋能劳动力转型与供需匹配机制 subgraph S1["劳动力市场痛点与机遇"] A1["超 4000 万失业 / 低薪人群"] A2["医疗、物流、技工等超 260 万职位空缺"] end subgraph S2["AI 介入的双重路径"] B1["赋能个体：个性化转型工具与技能培训"] B2["赋能体系：服务数据聚合与效果反馈优化"] end subgraph S3["最终社会价值"] C1["改变个体家庭生活轨迹"] C2["满足企业用工需求"] C3["促进整体经济增长"] end A1 --> B1 A2 --> B1 B1 --> B2 B2 --> C1 B2 --> C2 B2 --> C3 ``` AI 在加速并优化这一转型过程中能够扮演决定性角色，其落地主要体现为两大方向：其一是为处于转型期的个体提供全套支持工具与赋能手段；其二是沉淀并分析服务交付全流程的数据，从而大幅度改进劳动力服务体系的效率。在基础培训层面，Goodwill 已经直接支持了约 25,000 人接受重要 AI 工具与应用技能的培训。 - **4000 万+**：全美处于失业、不充分就业或低薪状态的人群总数。 - **200 万+**：医疗保健与物流领域的职位空缺数。 - **60 万+**：技术工种、建筑维护等领域的职位空缺数。 - **25,000 人**：Goodwill 已直接完成 AI 技能培训的人数规模。 --- ### [04:18](https://www.youtube.com/watch?v=dn75IsD4gD8&t=258s) 数字化与 AI 赋能的端到端职业辅导全流程在 Goodwill 服务的群体中，约 85% 至 86% 的人最高学历仅为高中或以下，许多人面临严峻的生活挑战，包括曾受司法系统处理（出狱人员）、长期失业等。当一名求职者走进 Goodwill，职业导师（Career Coach）需要为其提供从软技能、就业准备度评估到职业技能培训的全面支持。在传统模式下，导师需要手动让学员完成三项评估（生活现状、就业准备度、技术技能掌握情况），再结合本地劳动力市场数据定制路线图，并处理学员面临的各种衍生困境： - 如果学员面临住房不稳定，导师需要转介外部救助机构； - 如果学员心理压力过大，需要对接心理辅导； - 如果学员需要育儿支持或交通协助，必须多方协调社会网络； - 随后还要辅导学员完成数字技能课程、修改简历及模拟面试。 ```mermaid flowchart LR accTitle: AI 赋能的职业发展与支持全流程 subgraph Intake["评估与诊断"] A["三项综合评估\n(生活/就业准备/技能)"] --> B["AI 聚合分析\n匹配本地就业市场"] end subgraph WrapAround["综合服务与学习"] B --> C["第三方资源网络联动\n(住房 / 心理 / 托儿)"] B --> D["自适应 1 对 1 在线学习\n(数字与岗位技能)"] end subgraph Placement["就业与反馈"] C --> E["就业冲刺支持\n(简历/面试/岗位匹配)"] D --> E E --> F["雇主录用与长期跟踪"] end ``` AI 的引入能够重塑这一链条：AI 能在瞬间完成多项评估数据的整合并生成高度定制化的成长路线图；能够自动化对接第三方住房、心理健康与托儿服务网络并追踪使用情况；能够在学员进行线上课程学习时提供自适应的 1 对 1 个性化辅导。更重要的是，将职业导师从繁重的数据录入与行政事务中解放出来后，导师可以将全部精力投入到真正需要人类温度的深度共情与辅导中，从而从根本上提升整个劳动力支持系统的质量。 --- ### [08:07](https://www.youtube.com/watch?v=dn75IsD4gD8&t=487s) 当前行业生态现状与大规模数据聚合的未来潜力针对主持人关于“当前离这一理想蓝图还有多远”的提问，Steve Preston 指出，目前整个行业仍处于创新的“点状分布”阶段。市场上既有小微创新团队的单点突破，也有大型机构的重磅投入——例如 Coursera 投入巨资开发自适应学习工具；盖茨基金会（Gates Foundation）、鲍尔默基金会（Ballmer Group）以及 Stand Together 等大型慈善机构也在出资支持那些搭建跨领域连接系统的创新企业（如专门针对刑释人员重返社会的对接工具）。然而，目前各单点工具之间的底层“连接纽带”（Connective Tissue）尚未健全。 Goodwill 的核心优势在于其巨大的实体覆盖体量： - 每年触达服务超过 **200 万人**； - 每年为超过 **20 万人** 提供深度的“一对一”密集型就业个案管理服务。 Goodwill 目前正与外部创新力量共同研发和测试这些工具。而在未来，AI 最具突破性的应用在于**跨周期的大规模数据聚合**。过去虽然积累了局部经验，但整个行业缺乏深层的大数据支撑来明确回答：“面对特定困境特征的人群，究竟哪 4 到 5 项关键干预措施能真正撬动改变？” Steve Preston 将这一机制比作现代医学诊断：“就像医生面对特定病症，能根据持续输入的多维指标为你提供精准诊疗一样，劳动力支持系统在社会层面上也应具备这种精准干预能力。”以出狱人员为例，全美刑释人员在获释后 1 到 2 年内的重新犯罪率（Recidivism Rate）高达 30% 至 40%；而 Goodwill 依托精准研究与深度个案介入的司法矫正支持项目，成功将这一人群的重犯率压缩到了“中个位数”（即 4% - 6% 左右）的水平，充分证明了数据驱动与精准支持的巨大威力。 ### [12:05](https://www.youtube.com/watch?v=dn75IsD4gD8&t=725s) 数据与 AI：打通摆脱贫困与填补企业人才缺口的双向桥梁在特定的微观试点项目中，精准的协同支持已经带来了令人惊叹的极低再犯率（Recidivism Rate）。Steve Preston 强调，如果能将这种模式推广到 Goodwill 的所有业务领域，利用数据洞察与 AI 工具，就能为深陷贫困与困境的人群打造一条根本性的上升通道。这不仅是救助弱势群体，更是为急需劳动力的雇主输送关键人才、驱动经济增长的双赢之举，堪称当今社会“最完美的共生关系”（The most symbiotic thing in society）。 ```mermaid flowchart LR accTitle: 弱势群体通过 AI 与数据赋能实现社会共生循环 subgraph S1["赋能与转型"] A["弱势群体 / 服刑人员"] --> B["AI 赋能 + 数据精准支持"] B --> C["技能提升与生活重塑"] end subgraph S2["经济与社会回报"] C --> D["企业关键岗位人才"] D --> E["驱动经济与打破代际贫困"] end ``` - **双向赋能**：通过 AI 工具和高质量数据带来的深度理解，帮助个体走出困境，同时解决企业面临的劳动力短缺问题。 - **微观到宏观**：在特定支持项目中已验证的低再犯率成果，具备全面推广并重塑整个劳动力发展体系的潜力。 --- ### [13:02](https://www.youtube.com/watch?v=dn75IsD4gD8&t=782s) 对不同领域与会者的行动建议：打破固有偏见与协同创新面对由创业者、大中小企业管理者、个人贡献者及 AI 专家组成的多元化参会群体，主持人 Adam Mendler 询问大家应如何切身参与到这场变革中。Steve Preston 给出了针对性的行动建议： - **人力资源与人才管理者（HR & Talent Leaders）**：必须认识到“潜能与才华是均匀分布的，但机会并非如此”（Talent and potential is evenly distributed, opportunity is not）。企业需要转变思维，重新审视人才培养与选拔机制，主动与像 Goodwill 这样专注于非传统人才培育的机构合作。 - **IT 与咨询合作方**：在推进技术方案与咨询服务时，主动探讨并引入能赋能人才发展的 AI 工具，发掘技术在公益与职业培训中的释放潜力。 - **技术投资机构与企业**：积极与当地非营利组织（如 Goodwill）展开合作，深入一线场景寻找具体的技术赋能切入点，通过局部实践摸索出可行模式并加以推广。 --- ### [15:41](https://www.youtube.com/watch?v=dn75IsD4gD8&t=941s) 改变社会叙事：从投资回报率（ROI）视角重估人才支持 Steve Preston 呼吁彻底改变社会对弱势群体的固有认知。在 Goodwill 的日常工作中，无数原本深陷泥潭的人在短短数月内彻底扭转了人生轨迹，进入全美顶尖机构任职。要实现这种转变，社会必须确立两个信念：“坚信改变是可能的”（Believe it's possible）和“坚信这笔投入是值得的”（Believe the investment is worth it），并在此基础上打破阻碍前行的体制机制壁垒。曾长年从事投资银行业务的 Steve 从财务与社会综合 ROI 的角度算了一笔账： | 评估维度 | 放任现状的社会成本 | 主动投资与赋能的价值回报 | |:--- |:--- |:--- | | **直接财政支出** | 每名服刑人员每年超 40,000 美元的关押成本 | 较低的技术与培训投入即可帮助其自立 | | **家庭与社会影响** | 父母缺位带来的多代负面连锁反应、司法系统的空转损耗 | 形成全新的家庭成长轨迹，子女获得更好的发展前景 | | **经济效益** | 劳动力流失与社会福利负担加重 | 为企业输送急需人才，创造持续税收与经济增长 | Steve Preston 直言，这种投入产出比是“纽约或硅谷任何一位投资银行家或私募股权投资人随时都乐意接受的绝佳 ROI”。非营利组织在深切关怀服务对象的同时，整个社会与企业界也应认识到这种投资带来的巨大经济与社会价值，而 AI 正是加速这一进程的关键解锁工具。最后，主持人 Adam Mendler 向 Steve 的精彩分享表达了感谢，全场在掌声中圆满结束对谈。 ## Notable quotes > [00:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=4s) 行业讨论在不久之前还停留在技术本身“能做什么”，而如今核心议题已经演变为 AI 对人类个体、社会福祉以及人才培养路径意味着什么。 — Steve Preston > [02:04](https://www.youtube.com/watch?v=dn75IsD4gD8&t=124s) 如果能对处于困境中的人群进行针对性技能培训，不仅能彻底改变个人及其家庭的命运轨迹，还能为企业输送急需人才并促进经济发展。 — Steve Preston > [04:18](https://www.youtube.com/watch?v=dn75IsD4gD8&t=258s) 将职业导师从繁重的数据录入与行政事务中解放出来后，导师可以将全部精力投入到真正需要人类温度的深度共情与辅导中，从而从根本上提升整个劳动力支持系统的质量。 — Steve Preston > [08:07](https://www.youtube.com/watch?v=dn75IsD4gD8&t=487s) 就像医生面对特定病症，能根据持续输入的多维指标为你提供精准诊疗一样，劳动力支持系统在社会层面上也应具备这种精准干预能力。 — Steve Preston > [12:05](https://www.youtube.com/watch?v=dn75IsD4gD8&t=725s) 这不仅是救助弱势群体，更是为急需劳动力的雇主输送关键人才、驱动经济增长的双赢之举，堪称当今社会“最完美的共生关系”。 — Steve Preston > [13:02](https://www.youtube.com/watch?v=dn75IsD4gD8&t=782s) 潜能与才华是均匀分布的，但机会并非如此（Talent and potential is evenly distributed, opportunity is not）。 — Steve Preston > [15:41](https://www.youtube.com/watch?v=dn75IsD4gD8&t=941s) 坚信改变是可能的（Believe it's possible）和“坚信这笔投入是值得的”（Believe the investment is worth it）。 — Steve Preston > [15:41](https://www.youtube.com/watch?v=dn75IsD4gD8&t=941s) 这种投入产出比是“纽约或硅谷任何一位投资银行家或私募股权投资人随时都乐意接受的绝佳 ROI”。 — Steve Preston ## Mentioned Steve Preston, Adam Mendler, Goodwill Industries International, Thirty Minute Mentors, Ai4, 美国住房和城市发展部 (HUD), 美国中小型企业管理局 (SBA), Coursera, 盖茨基金会 (Gates Foundation), 鲍尔默基金会 (Ballmer Group), Stand Together ## Video description Adam Mendler - Creator & Host, Thirty Minute Mentors Steve Preston - CEO, Goodwill Industries Internationa As AI transforms industries, organizations, and jobs, what can individuals do to stay relevant, and what role should companies play in ensuring employees are AI literate? A recent report from Goodwill Industries International (GII) indicates that only 22 percent of Americans use AI for career training and upskilling, while 35 percent report not using AI professionally at all. Forty-nine percent of respondents said they cannot keep up with the pace at which AI is moving, and only 40% said their employers provide adequate AI training. As GII President and CEO, Steve Preston leads both the largest nonprofit workforce development provider in the country and the largest network of thrift stores. In this fireside chat, Steve will talk about the current AI gaps in the workforce, the pathways we need to create for individuals to bridge to the AI future, how Goodwill is using AI to improve operations, what companies should do to ensure a competitive workforce in the AI world, and how individuals can leverage this pivotal period to advance their own careers. [Source](https://www.youtube.com/watch?v=dn75IsD4gD8)

Source quotes

[0:04] Steve Preston: 行业讨论在不久之前还停留在技术本身“能做什么”，而如今核心议题已经演变为 AI 对人类个体、社会福祉以及人才培养路径意味着什么。 [2:04] Steve Preston: 如果能对处于困境中的人群进行针对性技能培训，不仅能彻底改变个人及其家庭的命运轨迹，还能为企业输送急需人才并促进经济发展。 [4:18] Steve Preston: 将职业导师从繁重的数据录入与行政事务中解放出来后，导师可以将全部精力投入到真正需要人类温度的深度共情与辅导中，从而从根本上提升整个劳动力支持系统的质量。 [8:07] Steve Preston: 就像医生面对特定病症，能根据持续输入的多维指标为你提供精准诊疗一样，劳动力支持系统在社会层面上也应具备这种精准干预能力。 [12:05] Steve Preston: 这不仅是救助弱势群体，更是为急需劳动力的雇主输送关键人才、驱动经济增长的双赢之举，堪称当今社会“最完美的共生关系”。 [13:02] Steve Preston: 潜能与才华是均匀分布的，但机会并非如此（Talent and potential is evenly distributed, opportunity is not）。 [15:41] Steve Preston: 坚信改变是可能的（Believe it's possible）和“坚信这笔投入是值得的”（Believe the investment is worth it）。 [15:41] Steve Preston: 这种投入产出比是“纽约或硅谷任何一位投资银行家或私募股权投资人随时都乐意接受的绝佳 ROI”。

Source quotes

[0:04] Steve Preston: what it means for people, what it means for humanity, what it means for how we develop ourselves. [12:05] Steve Preston: astoundingly low recidivism rates [12:05] Steve Preston: the most symbiotic thing you can imagine [13:02] Steve Preston: talent and potential is evenly distributed; opportunity is not [15:40] Steve Preston: some of the most storied organizations in our country [15:40] Steve Preston: any investment banker or private equity person in New York or in the valley would take any time any day of the week.

Original transcript

[0:04] I'd like to introduce Steve Preston. Steve is the leader of the largest workforce [0:10] development organization in America. Steve has an incredible and incredibly diverse background. [0:18] Steve has served at the highest levels in the private sector, in the public sector, leading [0:24] private companies, leading public companies, serving as a cabinet secretary, serving as the [0:31] head of the SBA. Steve, thank you for being here. Yeah, thanks for having me. Nice intro. Thanks. [0:38] You have a very unique background. I would imagine you have a unique perspective on AI. How do you [0:46] view AI today? How do you utilize AI today? Yeah. Well, uh, obviously one of the great things about [0:52] this conference is AI is super multi-layered. And I think not that long ago, we were talking [0:58] about sort of, you know, technologically what it could do. But now we're talking about what [1:02] it means for people, what it means for humanity, what it means for how we develop ourselves. Um, [1:08] and so right now we are very focused on on uh on one very big issue. We've got we we kind of [1:15] deal with everything because we've got 140,000 employees and a big retail network and a big [1:20] workforce development network. But really the core of what we're focused on right now is what does [1:24] it mean for people uh employers and society. um and that moves um partially that involves [1:32] tooling people with AI skills and AI AI sort of enabling uh competencies but I think much more [1:38] broadly it relates to how do we think about this massive workforce transition we're going through [1:44] which is not just AI related um and employ AI to help us get to a much much better place for people [1:52] in our society for our economy and for employers and that's a much bigger question and that's uh [1:59] and and and that's a very big part of the nut that we're focused on. You posed the question, [2:04] how do you answer it? [laughter] So, um I think a lot of what we're focusing on here today uh are [2:11] AI skills and AI implementation in companies or other employers. Much bigger than that is we have [2:20] um many other transitions happening in our world right now. We have an aging population. We have [2:25] changing demographics in other ways. And we have a very significant portion of our workforce that [2:32] is really not in a great place. Over 40 million people in our country are unemployed, undermployed [2:39] or earning very low wages. At the same time, if you look at sort of the uh the employment side, [2:46] we have nearly two million openings in health care, uh in logistics, well over 600,000 in in the [2:53] trades, in other areas like building maintenance and places like that. And it doesn't look like [2:59] it's going to get a lot better unless we can skill those people. Many of the people on um who [3:06] are experiencing great challenges in life could be trained and toolled for those jobs which are [3:13] terrific jobs. And if we can do that, we solve a lot of problems, right? We help individuals [3:19] in their families in tough places achieve a much better life, fundamentally different trajectory. [3:24] We help employers get talent they need to compete. We help our economy grow. We solve problems in our [3:31] communities. And AI can be an incredible tool to manage that transition much more effectively [3:38] than we can today. And uh and it can do so in two ways. I would say broadly number one providing the [3:46] tools and capabilities to support people in that transition and getting the data on how we're doing [3:53] so that we can dramatically improve the way we can deliver those services. Um and so those are two [3:59] areas. I can flesh those out some more, but those are those are two big areas we're beginning to [4:04] focus on beyond training people in AI skills. And I would just mention we've uh we've directly uh [4:11] supported training about 25,000 people already uh in in important AI tools and capabilities. Steve, [4:18] can you dive into those two areas? Yeah. So, so think about let let's start with the individual [4:23] uh for a minute. So, many of the pe the 85 86% of the people we serve have a high school degree or [4:28] less. Many of them face great challenges in life. Many of them are justice impacted individuals, [4:34] people who've had other kinds of challenges in life, um maybe have been unemployed quite [4:39] a bit. So, somebody comes into a Goodwill and what we do is we work with that individual to [4:45] do whatever we can to help them get on their feet, to help them get job skills, soft skills, [4:52] job readiness, other kind of help that they need, and eventually move them into employment in an [4:59] onto a much better trajectory in life. What does that involve? It requires a lot of individual [5:05] support, a lot of data, a lot of training, a lot of connective tissue. So, I come into Goodwill, [5:12] you're my you're my career coach, right? Um and you say, "Steve, I need you to take three [5:18] assessments. Uh sort of where you are in life, your workforce readiness, are you really ready for [5:24] the workforce or for the kinds of jobs that we're looking at? what kind of technical skills do you [5:29] have? I get all of those assessments and I make sense of those and and you say to me, well, Steve, [5:35] these are the jobs. Here's the local job market data. Here's what's available in your in your area [5:40] and but here are kind of the challenges I'm seeing for you and the things we need to work on. And so, [5:45] let's talk about a journey for you that gets you from point A to point B. And I'm going to talk to [5:52] you about gee, you're housing insecure. I need to refer you someplace. can you get over there? Um, [5:57] you need sort of some foundational work readiness skills. I want to connect you with a counselor [6:04] over here to support you with other challenges you might have. And throughout that journey, [6:09] now you're going to take a class and I'm going to help you get through a digital skills class. [6:13] Maybe think about the following. Think about the aggregation of that information and turning it [6:20] into an individual journey which can happen like that with AI. Think about actually the [6:28] connection with third-party providers that can give you housing support or that can provide you [6:34] with counseling or can provide you with child care. Uh think about the the ability to build [6:40] networks to connect with those players and track that individual and their and their usage of it. [6:47] Now think about the actual learning journey and somebody who's taking a class much of which would [6:53] be online and think about one on-one personalized learning to support that individual in their [6:59] unique journey and then think about supporting that person with all of that to ensure that when [7:06] they're ready for that point they are getting the support they need to find the job in their local [7:10] job market the resume the interview skills etc etc we do that um that's what we do today. We do [7:19] all of that today. But if if if you're my career coach, think about what that means for you to have [7:24] that available without you doing all of that work. Think about how your time will be spent providing [7:31] much more individualized support coaching. You'll have all of those uh decision-making tools. You [7:38] will be connected with providers in your area much more effectively. And you will be able to [7:43] ensure employers when they hire when they hire me or whatever individual you're working with [7:48] that that person has gotten the kind of support they need. All of those components are being [7:54] developed in pockets in different places and that could fundamentally transform the quality [8:00] of support we give to people in a system that doesn't really work very well today. [8:07] Steve, you laid out the north star, the vision. Yeah. And it sounds like we're getting there, [8:14] but not quite there. How close are we to getting there? And what is required to ultimately allow [8:21] us to get there? Yeah. So um first of all uh it what I would say what we see is a lot of things [8:30] happening in little pockets and it's sort we're sort of at that place in the development where [8:35] you have um small innovators in many cases uh develop some of these tools. Um we we recently saw [8:43] um a very large company indeed make a significant invest excuse me Corsera make a significant [8:49] investment to divide to um develop sort of that adaptive learning. So we see kind of you know big [8:56] companies we see small innovators um you know uh a number of big foundations uh Gates Balmer [9:03] stand together and others have just invented uh invested in entity that uh invest in companies [9:09] that do a lot of the connective tissue in very specific places some of it for people coming [9:14] out of incarceration some of it for other things so you see all of these activities happening in [9:19] different places I think what we're going to see over time is quality tools kind of be developed, [9:26] adopted, tested, and over time go into usage. The connective tissue across those tools are [9:33] not in place. And it doesn't really seem right now that there's a lot of we're at that point [9:37] in the development, but what we're trying to do at Goodwill is work with people out there [9:43] because we we touch two million people a year. Okay? We provide over 200,000 people a year with [9:51] intensive job services. And that means an individual is assigned to that, you know, [9:56] that person in need and and actually works with them on the journey I just described. So we have a [10:02] massive population that we support and what we're trying to do is begin to work with those people to [10:07] to begin to test and and sort of co-develop many of those tools. The other thing that could be just [10:15] groundbreaking for all of us um which is very also a very difficult thing that AI could help us break [10:21] through is the aggregation of data of what's happening with all those individuals so we can [10:29] see what works. Now we know through experience what works in many cases and we know locally [10:36] we'll have data but when you think about that whole you know kind of chain I described for you [10:44] um there isn't a lot of data that says these are the four or five things that really matter [10:53] for a person in a particular condition. This is what will really move the needle. And if we can [10:58] begin aggregating that on a much broader level, we have um much better information for research, [11:06] for policy makers, for practitioners like us who perform that ser those services locally to be able [11:13] to say when com somebody when I get those decision support tools, these are the things the data show [11:19] work. This is how you need to pivot based on what you're seeing in that individual's life. I mean, [11:24] think about think about a doctor, you know, when you're going and you're you're you're being [11:28] diagnosed with a particular condition and that doctor is kind of working with you through the [11:32] journey and more information comes in and they're able to guide you. We should have that same [11:36] ability as a society on a much bigger level. Right now, just an example, you know, when somebody when [11:43] when we see people coming out of incarceration, the recidivism rates are massive. You know, [11:49] 30 40% the first year or two. Um we've seen in you know our core our core uh uh um justice involved [11:58] program mid-s singledigit recidivism rates because we intensively study that what works we work [12:05] together to make sure that we're delivering support we need and we have astoundingly low [12:10] recetivism rates that is in a very sort that's in a name microposm it's in a very specific program [12:16] if we understood that across all of the different kind of work we do here and we're able to provide [12:24] that um to people you know providing these [clears throat] services we would see a fundamentally [12:30] different pathway for people out of poverty out of very difficult situations and that pathway [12:37] would be providing critical talent for employers who need that talent to fuel our economy and our [12:44] growth. I mean it's the most symbiotic thing you can imagine you know in society right now. [12:49] Um but we need to build those bridges and in in between AI tools and better understanding [12:54] through better data. There's just a massive unlock that I think is available for all of us at this [13:02] conference. We have people who are at all levels. We have entrepreneurs. We have leaders of large [13:09] organizations, small organizations, individual contributors, people who are AI experts, people [13:14] who are interested in learning about AI. So it's a diverse audience. What advice would you give to [13:20] the different members of this room, the different members of this conference on how to get engaged, [13:29] how to get involved in tackling everything that you just shared? Yeah. So I mean for you know [13:35] those of you who are uh you know kind of in the HR space or um looking broadly at talent issues [13:44] um I would strongly encourage you number one to understand there is a massive untapped talent [13:51] pool in our country right you know uh talent and potential is evenly distributed um opportunity is [14:00] not and in many cases it's opportunity It's access and because of some of the challenges people have [14:07] had in life, it is thinking differently about how we train, develop and nurture people into [14:13] these roles and and and so if you're in those roles, begin to work with your partners to look [14:21] differently um at talent and to go to places that develop talent that are different like [14:26] us to to begin to understand, you know, h how you could fill those roles in that process. [14:33] If you are working with people on these challenges, if you are working with people in the [14:38] um uh in you know uh because many of you work with consultants, many of you have IT uh relationships, [14:45] encouraging them, asking them about the tools, um helping them understand there's an unlock. [14:51] If you're at a if you're at a, you know, an organization that is actually investing in or [14:56] thinking about these types of tools, absolutely. um you know work with people like Goodwill, [15:02] work with other people in your uh local nonprofits who have these challenges to to to understand [15:09] where those unlocks are. We're seeing it in little kind of, you know, in microcosms everywhere. And [15:15] I think the more that we do that, the more that we're going to learn what works, the more that [15:20] we're going to get this conversation going. Um, and you know, I I just have to say this conference [15:25] I'm so um uh appreciative that there's so many people here that are focused on talent issues [15:32] um because they really are beginning to think about these things. But um yeah, [15:36] it it it's going to it's going to have to happen a lot of different places. Steve, is there anything [15:41] else you'd like to share? Um I would just, you know, uh yeah, I've got a lot I could share. I've [15:49] got a whole lot I could share. I think the other thing I would share and this is not a very AI [15:54] thing. Uh but the thing I am really hoping so much is that part of what we what we will accomplish [16:04] here is we will change the narrative on what's possible for people in our society. You know, [16:11] I mentioned the recidivism rate uh for people who've been incarcerated. Um, you know, [16:16] we drive through communities all the time in our cities and see people and just wonder, you know, [16:21] uh, gosh, is this ever going to get better or what can happen? Um, we have stories about people, [16:27] um, based on what we've read or our experiences. And what I would tell you is what we see every [16:34] day at Goodwill is people that you would never have thought would have moved from [16:40] a terribly difficult situation in a series of months. fundamentally changing their lives, [16:46] um, fundamentally overcoming challenges and moving into some of the most storied organizations in our [16:54] country with uh, terrific roles with terrific futures and completely transformed lives. And [17:02] and and you know, I often tell people number one, we we have to believe it's possible. [17:10] We have to believe the investment is worth it. And if we believe those two things, we have to bust [17:16] through the systems that inhibit our ability to move forward. And I would tell you, okay, [17:22] getting back to, you know, my my incarceration example, um sometimes we say, gosh, how much money [17:29] is that going to take or what is it going to, you know, um is it really worth it? Let me tell you, [17:34] um, when you look at somebody who is incarcerated for $40,000 plus a year, when you think of their [17:41] family, uh, and and the support that they need, when you think about the multi-generational [17:46] impact of not having that parent in the home, when you think about the enormous [17:51] cost of people churning into the just back and forth in the justice system, and then you say, [17:58] okay, well, what would it take to invest this individual either technologically or otherwise? [18:03] to help them move forward in life and become somebody who can take care of themselves and [18:09] their families who have a fundamentally different trajectory in life whose children have a better [18:13] pathway forward. That is I mean I used to I was an investment banker for a long time were [18:19] formed. Um that is an ROI any investment banker or private equity person in New York or in the [18:26] valley would take any time any day any day of the week. the the the cost the the financial cost and [18:33] the cost to society and families of what we abide in our country versus the value of changing that [18:41] is an incredible return on investment. And so um the the thing the thing I I work you know [18:49] want very much that people understand is yes those of us who do this work care deeply about [18:55] the individuals we serve care deeply about their lives but when you think of the financial ROI to [19:01] society when you think about the broader ROI just the to communities and and and the value that all [19:07] of us believe in the return is massive and every one of us should be deeply committed to to you [19:14] know advancing these opportunities for people in society and for employers. And once again, [19:20] you know, I think AI could be an incredible unlock for those of us doing this work. Steve, [19:27] thank you for all the great advice. Thank you for being here. And thank you [applause]