Future Of Work

🚨 BREAKING: Taiwan enacted its basic law on AI, which includes, among other innovative provisions, detailed AI governance principles and LABOR RIGHTS for humans who lose their jobs due to AI. Other countries should take note: According to the law's third article, the research and application of AI in Taiwan should adhere to the following principles (read them carefully!): 1. Sustainability: It should consider mental health, social equity, and environmental sustainability, reducing potential health risks or digital disparities, and enabling the public to adapt to the changes brought about by AI. 2. Human Autonomy: It should support human autonomy, respect fundamental human rights and cultural values such as the right to personality, allow for human oversight, and implement a people-centered approach that respects the rule of law, human rights, and democratic values. 3. Privacy Protection and Data Governance: It should respect the privacy and autonomy of personal data, adopt the principle of data minimization, and avoid the risk of data leakage. 4. Security: Cybersecurity measures should be established throughout the research and application of AI to prevent security threats and attacks, ensuring the robustness and security of the system. 5. Transparency and Explainability: AI outputs should be appropriately disclosed or labeled to facilitate risk assessment and understanding of their impact on relevant rights, thereby enhancing the trustworthiness of AI. 6. Fairness: AI research and application should avoid risks such as system bias and discrimination, and should not result in discrimination against specific groups. 7. Accountability: Traceability should be maintained, and different roles in AI research and application should bear corresponding responsibilities, including internal governance responsibilities and external social responsibilities. For those familiar with the EU AI Act, the way the principles above are framed is more direct and comprehensive than the European framework. As I wrote a few times before, the EU missed an opportunity to be more explicit and broad when protecting fundamental rights in the context of AI development and deployment (which could help set a stronger regulatory precedent). Another interesting provision is Article 12, focused on labor rights. It says that, in response to the development of AI, the government must address skill gaps and ensure workers' occupational safety, health, and labor rights, including providing employment assistance to those unemployed due to AI, based on their work abilities. To my knowledge, this is the first AI law that expressly foresees labor rights for those who lose their jobs due to AI. Well done, Taiwan! - 👉 To learn more about recent AI governance developments, join my newsletter's 90,000+ subscribers (below). 👉 To upskill and advance your career, join the 28th cohort of my AI Governance training in March (link below).

We’ve all heard about AI’s potential to boost productivity. But what truly matters to me is whether it’s making work better for the people who show up every day. At Cisco, our People Intelligence team, in collaboration with IT, has been exploring this very topic, and the findings are fascinating. Here are five key insights from our research that leaders should take seriously: 1. Leaders are key to adoption. At Cisco, employees are 2x more likely to use AI if their direct leader uses it. 2. Generic AI training doesn’t work. Role-specific, practical training accelerates AI use. 3. Confidence gaps exist among senior leaders. Directors at Cisco often feel less confident with AI than mid-level employees, underscoring the need for tailored support at all levels. 4. Employee autonomy fuels adoption. Hybrid work environments are powerful accelerators for AI adoption, while mandates can hinder it. Employees who voluntarily go to the office are more likely to use AI, while those who are required to work on-site have lower adoption. 5. AI use is linked to employee well-being, but the relationship is complex, with both benefits and trade-offs that require thoughtful navigation. This is just the beginning. Next, we’re looking at how AI is transforming the way teams operate. For now, one thing is clear, employees who use AI aren’t just more productive. They’re also more engaged, better aligned with company strategy, and empowered to focus on meaningful work. #AIAdoption #EmployeeExperience #FutureOfWork

A Return To Office mandate is a funny thing. A trade-off of lower workforce productivity, morale, retention, engagement, and trust in exchange for...managers feeling more in control. It's more a sign of insecurity and incompetence than sound decision-making. The fact that 80% of executives who have pushed for RTO mandates have later regretted their decision only makes the point further, and yet every few months more leaders line up to pad this statistic. In case your leaders have forgotten, return to office mandates are associated with: 🔻 16% lower intent to stay among the highest-performing employees (Gartner) 🔻 10% less trust, psychological safety, and relationship quality between workers and their managers (Great Place to Work) 🔻 22% of employees from marginalized groups becoming more likely to search for new jobs (Greenhouse) 🔻 No significant change in financial performance while guaranteeing damage to employee satisfaction (Ding and Ma, 2024) The thing is, we KNOW how to do hybrid work well at this point. 🎯 Allow teams to decide on in-person expectations, and hold people accountable to it—high flexibility; high accountability. 🎯 Make in-person time unique and valuable, with brainstorming, events, and culture-building activities—not video calls all day in the office. 🎯 Value outcomes, not appearances, of productivity—reward those who get their work done regardless of where they do it. 🎯 Train inclusive managers, not micromanagers—build in them the skills and confidence to lead with trust rather than fear and insecurity. Leaders that fly in the face of all this data to insist that workers return to office "OR ELSE" communicate one thing: they are the kinds of leaders that place their own egos and comfort above their shareholders and employees alike. Faced with the very real test of how to design the hybrid workforce of the future, these leaders chose to throw a tantrum in their bid to return to the past, and their organizations will suffer for it. The leaders that will thrive in this time? Those that are willing to do the work. Those that are willing to listen to their workforce, skill up to meet new needs, and claim their rewards in the form of the best talent, higher productivity, and the highest level of worker loyalty and trust. Will that be you?

Louder for the people at the back 🎤 Many organisations today seem to have shifted from being institutions that develop great talent to those that primarily seek ready-made talent. This trend overlooks the immense value of individuals who, despite lacking experience, possess a great attitude, commitment, and a team-oriented mindset. These qualities often outweigh the drawbacks of hiring experienced individuals with a fixed and toxic mindset. The best organisations attract talent with their best years ahead of them, focusing on potential rather than past achievements. Let’s be clear this is more about mindset and willingness to learn and unlearn as apposed to age. To realise the incredible potential return, organisations must commit to creating an environment where continuous development is possible. This requires a multi-faceted approach: 1. Robust Training Programmes: Employers should invest in comprehensive training programmes that equip employees with the necessary skills for their roles. This includes on-the-job training, mentorship programmes, online courses, and workshops. 2. Redefining Hiring Criteria: Organisations should revise their hiring criteria to focus more on candidates’ potential and willingness to learn rather than solely on prior experience or formal qualifications. Behavioural interviews, aptitude tests, and probationary periods can help assess a candidate's ability to learn and adapt. 3. Partnerships with Educational Institutions: Companies can collaborate with educational institutions to design curricula that align with industry needs. Apprenticeship programmes, internships, and cooperative education can bridge the gap between academic learning and practical job skills. 4. Lifelong Learning Culture: Encouraging a culture of lifelong learning within organisations is crucial. Employers should provide ongoing education opportunities and support for professional development. This includes continuous skills assessment and access to resources for upskilling and reskilling. 5. Inclusive Recruitment Practices: Employers should implement inclusive recruitment practices that remove biases and barriers. Blind recruitment, diversity quotas, and targeted outreach programmes can help ensure that diverse candidates are given a fair chance. By implementing these measures, organisations can develop a workforce that is adaptable, innovative, and resilient, ensuring sustainable success and growth.

Just out in Harvard Business Review, summary of the Hybrid Experiment results and lessons on how to make hybrid succeed. Experiment: randomize 1600 graduate employees in marketing, finance, accounting and engineering at Trip.com into 5-days a week in office, or 3-days a week in office and 2-days a week WFH. Analyzed 2 years of data. Two key results A) Hybrid and fully-in-office showed no differences in productivity, performance review grade, promotion, learning or innovation. B) Hybrid had a higher satisfaction rate, and 35% lower attrition. Quit-rate reductions were largest for female employees. Four managerial lessons 1) Hybrid needs a strong performance management system so managers don’t need to hover over employees at their desks to check their progress. Trip.com had an extensive performance review process every six months. 2) Coordinate in-office days at the team or company level. Schedule clarity prevents the frustration of coming to an empty office only to participate in Zoom calls. Trip.com coordinated WFH on Wednesday and Friday. 3) Having leadership buy-in is critical (as with most management practices). Trip.com’s CEO and C-suite all support the hybrid policy. 4) A/B test new policies (as well as products) if possible. Often new policies turn out to be unexpectedly profitable. Trip.com made millions of dollars more profits from hybrid by cutting expensive turnover.

Last week, I described four design patterns for AI agentic workflows that I believe will drive significant progress: Reflection, Tool use, Planning and Multi-agent collaboration. Instead of having an LLM generate its final output directly, an agentic workflow prompts the LLM multiple times, giving it opportunities to build step by step to higher-quality output. Here, I'd like to discuss Reflection. It's relatively quick to implement, and I've seen it lead to surprising performance gains. You may have had the experience of prompting ChatGPT/Claude/Gemini, receiving unsatisfactory output, delivering critical feedback to help the LLM improve its response, and then getting a better response. What if you automate the step of delivering critical feedback, so the model automatically criticizes its own output and improves its response? This is the crux of Reflection. Take the task of asking an LLM to write code. We can prompt it to generate the desired code directly to carry out some task X. Then, we can prompt it to reflect on its own output, perhaps as follows: Here’s code intended for task X: [previously generated code] Check the code carefully for correctness, style, and efficiency, and give constructive criticism for how to improve it. Sometimes this causes the LLM to spot problems and come up with constructive suggestions. Next, we can prompt the LLM with context including (i) the previously generated code and (ii) the constructive feedback, and ask it to use the feedback to rewrite the code. This can lead to a better response. Repeating the criticism/rewrite process might yield further improvements. This self-reflection process allows the LLM to spot gaps and improve its output on a variety of tasks including producing code, writing text, and answering questions. And we can go beyond self-reflection by giving the LLM tools that help evaluate its output; for example, running its code through a few unit tests to check whether it generates correct results on test cases or searching the web to double-check text output. Then it can reflect on any errors it found and come up with ideas for improvement. Further, we can implement Reflection using a multi-agent framework. I've found it convenient to create two agents, one prompted to generate good outputs and the other prompted to give constructive criticism of the first agent's output. The resulting discussion between the two agents leads to improved responses. Reflection is a relatively basic type of agentic workflow, but I've been delighted by how much it improved my applications’ results. If you’re interested in learning more about reflection, I recommend: - Self-Refine: Iterative Refinement with Self-Feedback, by Madaan et al. (2023) - Reflexion: Language Agents with Verbal Reinforcement Learning, by Shinn et al. (2023) - CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing, by Gou et al. (2024) [Original text: https://lnkd.in/g4bTuWtU ]

We spend a lot of time designing organizations. Structures. Reporting lines. Governance.   And yet, when performance slows down or transformations stall, the issue is rarely the formal design.   It is the network beneath it. Who people trust. Where information really flows. Who connects teams and where friction quietly builds.   In every organization, there is an “invisible layer” that determines how things actually get done.   In my experience, understanding this layer is one of the most powerful and most underused levers to improve performance.   Over the coming weeks, we will run an Organizational Network Analysis (ONA) in parts of our organization.   This is not about evaluating individuals. It is about understanding how the system works so that we can remove friction, strengthen collaboration, and operate at our full potential.   If you are interested in the thinking behind this, one of the most practical articles on the topic is in the comments.    When you make the invisible visible, performance follows.

Thank you, Google. You just open-sourced a single CLI for all of Google Workspace and it's built for both humans and AI agents. npm install -g @googleworkspace/cli What it does: → One command-line tool for Drive, Gmail, Calendar, Sheets, Docs, and every Workspace API → Zero boilerplate. Structured JSON output. Auto-pagination. → Reads Google's Discovery Service at runtime — when Google adds a new API endpoint, the CLI picks it up automatically → Ships with 100+ Agent Skills so your LLM can manage Workspace without custom tooling → Built-in MCP server for Claude Desktop, Gemini CLI, VS Code, and any MCP-compatible client → Model Armor integration to scan responses for prompt injection before they reach your agent This is a big deal for anyone building AI agents that interact with Google Workspace (everyone?) No more writing custom API wrappers. No more maintaining brittle integrations. One tool. Every service. Structured output ready for agents. The repo is Apache-2.0 licensed and under active development.

For years, companies have ramped up the surveillance of their employees, often without the employees' knowledge (or consent): monitoring what websites they accessed, their phone calls, their keystrokes, the speed with which they drove, and numerous other things. As this piece points out, the advent of AI makes actually analyzing all the tracking data much easier and less expensive, so the surveillance industry is about to get even bigger. In the U.S., when employees go to work, they give up their rights--to speak, to privacy, to criticize their companies and their decisions, and so many other things. As this piece also notes and as social science research has frequently demonstrated, job autonomy is an important determinant of motivation. With people under ever more monitoring, it is little wonder that job satisfaction and engagement, and for that matter trust in organizations, has fallen. When will companies learn to value human intelligence and motivation as much as they value the artificial variety? #motivation #surveillance #bossware #management #leadership #jobautonomy

𝗜𝘀 𝗯𝗲𝗶𝗻𝗴 𝗽𝗿𝗲𝘁𝘁𝘆 𝗴𝗼𝗶𝗻𝗴 𝘁𝗼 𝗴𝗲𝘁 𝘆𝗼𝘂 𝘁𝗵𝗮𝘁 𝗷𝗼𝗯? — 𝘴𝘵𝘶𝘥𝘺 𝘴𝘢𝘺𝘴 𝘮𝘢𝘺𝘣𝘦. We often talk about meritocracy in hiring — but research keeps reminding us how easily optics overshadow objectivity. 📊 A 2024 study by Harvard Business Review found that 𝗼𝘃𝗲𝗿 𝟱𝟬% 𝗼𝗳 𝗺𝗮𝗻𝗮𝗴𝗲𝗿𝘀 𝘀𝘂𝗯𝗰𝗼𝗻𝘀𝗰𝗶𝗼𝘂𝘀𝗹𝘆 𝗳𝗮𝗰𝘁𝗼𝗿 𝗶𝗻 𝗮𝗽𝗽𝗲𝗮𝗿𝗮𝗻𝗰𝗲 𝘄𝗵𝗲𝗻 𝗺𝗮𝗸𝗶𝗻𝗴 𝗵𝗶𝗿𝗶𝗻𝗴 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻𝘀. Similarly, research in the 𝘑𝘰𝘶𝘳𝘯𝘢𝘭 𝘰𝘧 𝘈𝘱𝘱𝘭𝘪𝘦𝘥 𝘗𝘴𝘺𝘤𝘩𝘰𝘭𝘰𝘨𝘺 𝘢𝘯𝘥 𝘗𝘴𝘺𝘤𝘩𝘰𝘭𝘰𝘨𝘪𝘤𝘢𝘭 𝘚𝘤𝘪𝘦𝘯𝘤𝘦 confirms that 𝗮𝘁𝘁𝗿𝗮𝗰𝘁𝗶𝘃𝗲 𝗰𝗮𝗻𝗱𝗶𝗱𝗮𝘁𝗲𝘀 𝗮𝗿𝗲 𝗺𝗼𝗿𝗲 𝗹𝗶𝗸𝗲𝗹𝘆 𝘁𝗼 𝗯𝗲 𝗿𝗮𝘁𝗲𝗱 𝗮𝘀 𝗰𝗼𝗺𝗽𝗲𝘁𝗲𝗻𝘁, 𝗰𝗼𝗻𝗳𝗶𝗱𝗲𝗻𝘁, 𝗮𝗻𝗱 𝗾𝘂𝗮𝗹𝗶𝗳𝗶𝗲𝗱 — even when their résumés are identical to less “polished” counterparts. This isn’t vanity; it’s psychology. It’s called the “𝗵𝗮𝗹𝗼 𝗲𝗳𝗳𝗲𝗰𝘁” — a cognitive bias where one positive trait (like appearance or confidence) spills over to how we judge unrelated qualities (like intelligence or leadership). And it’s costly. Because every time we let surface cues dictate selection, we risk 𝗺𝗶𝘀𝘀𝗶𝗻𝗴 𝘁𝗵𝗲 𝗾𝘂𝗶𝗲𝘁 𝗯𝗿𝗶𝗹𝗹𝗶𝗮𝗻𝗰𝗲 that doesn’t advertise itself well. The solution isn’t to ignore presentation — it’s to balance perception with structure: • 𝗨𝘀𝗲 𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲𝗱 𝗶𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄𝘀 and 𝘀𝗸𝗶𝗹𝗹-𝗯𝗮𝘀𝗲𝗱 𝗲𝘃𝗮𝗹𝘂𝗮𝘁𝗶𝗼𝗻𝘀. • Involve 𝗱𝗶𝘃𝗲𝗿𝘀𝗲 𝗶𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄 𝗽𝗮𝗻𝗲𝗹𝘀 to reduce individual bias. • Train leaders to recognize 𝗵𝗮𝗹𝗼 𝗮𝗻𝗱 𝗵𝗼𝗿𝗻 𝗲𝗳𝗳𝗲𝗰𝘁𝘀 — before they unconsciously act on them. 💬 𝘐𝘯 𝘭𝘦𝘢𝘥𝘦𝘳𝘴𝘩𝘪𝘱, 𝘴𝘦𝘦𝘪𝘯𝘨 𝘤𝘭𝘦𝘢𝘳𝘭𝘺 𝘣𝘦𝘨𝘪𝘯𝘴 𝘸𝘪𝘵𝘩 𝘴𝘦𝘦𝘪𝘯𝘨 𝘰𝘶𝘳 𝘰𝘸𝘯 𝘣𝘭𝘪𝘯𝘥 𝘴𝘱𝘰𝘵𝘴. #HiringBias #OrganizationalPsychology #Leadership #UnconsciousBias #DEI #FutureOfWork