Sohrab Salimi

Here is a summary of the ten lessons... hope you enjoyed it.

Uber let 23-year-olds launch entire cities. But they couldn't skip one meeting. Travis Kalanick explains how Uber managed its early expansion. Local teams had enormous freedom. But before launching, they needed his approval on pricing. That single decision required them to understand local regulations, driver costs, and how the city actually worked. Some early launches took hours of discussion. By around city 20, the pricing call was down to five minutes. Because every problem they solved became part of the playbook. That's a very different approach from managing people through endless rules and approvals. Give teams room to make decisions. Be clear about the few things that really matter. And make sure the next team doesn't have to solve the same problems again. What I find especially important is that Kalanick didn't keep himself at the center forever. Once the process worked without him, he stepped away. A leader's involvement should become less necessary as the organization learns. Source: Travis Kalanick, Excellence Is the Capacity To Take Pain, David Senra Podcast #Leadership #Autonomy #Scaling

Day 1 of AI for Leaders in Istanbul. 🇹🇷 One of the things I enjoy most about working with different organizations is seeing how differently they approach the same technology. And every workshop gives me something new to think about. Looking forward to continuing the conversation tomorrow. For those of you leading teams or organizations: What's your biggest challenge with AI right now? I'd love to hear what's happening on your side.

Why do we talk about AI's future as if we have no say in it? When Fei-Fei Li is asked about Geoffrey Hinton's warning that AI could lead to human extinction, she challenges the assumption behind the question. What are humans doing between now and then? We don't know how powerful AI will become. But companies are already deciding where to deploy it, what to automate, and how much control to give it. And here's what concerns me. How many organizations are moving ahead with AI without deciding who is accountable when it makes a mistake? Who can override it? Who decides when to switch it off? Those questions belong in the room before the technology goes live, not after something goes wrong. Source: Fei-Fei Li, The Mishal Husain Show, Bloomberg #AI #Leadership #AIResponsibility

Your job in a meeting isn't to sell your idea. It's to find out if it's wrong. Jeff Bezos explains why meetings at Amazon and Blue Origin start with 30 minutes of silent reading. Everyone reads the same six-page memo before the discussion begins. But his criticism of PowerPoint is what really matters here. Slides are great for persuasion. And that's precisely why they can be a problem when you're trying to make a good decision. It's easy to walk into a meeting prepared to defend your proposal. Much harder to walk in genuinely willing to change your mind. A good memo doesn't guarantee better decisions. But writing one forces you to work through your reasoning before asking others to agree with it. I don't care whether a team uses slides or memos. I care whether people feel free to challenge the idea, including the person who brought it into the room. Source: Jeff Bezos, Lex Fridman Podcast #405 #Leadership #DecisionMaking #Meetings

Silicon Valley may be inviting the regulation it fears most. Brian Chesky says something I wish more tech leaders would acknowledge. If AI companies resist regulation, they shouldn't be surprised when something goes wrong and governments respond with much stricter rules. But he also challenges policymakers. How can you write sensible rules for technology you don't understand? I don't want AI companies writing their own rules. But I don't want politicians writing them in isolation either. What worries me is how little these two sides seem to trust each other. And how rarely they appear willing to work through their disagreements. We need people who understand the technology and people who understand its impact on society making decisions together. I believe AI has enormous potential. But that doesn't give the companies building it a free pass. If you want the freedom to shape the future, you have to accept responsibility for what you might get wrong along the way. Source: Brian Chesky, Co-founder & CEO of Airbnb, Semafor's The CEO Signal #AI #Leadership #AIRegulation

If testing an idea takes less time than surveying people about it, why are we still surveying? Scott Kirby, CEO of United Airlines, recently launched a Starlink tracker without asking customers first. He talks to passengers, has a sense of what they might find useful, and prefers to see how they actually respond rather than spend weeks predicting it. Especially now that AI makes it possible to build and test ideas much faster than before. And I think a lot of companies are still operating as if every experiment requires a business case, three meetings, and management approval. Sometimes the experiment itself is the fastest way to get the evidence you're looking for. Of course, that only works when the risks are manageable and you can easily reverse the decision. But for small experiments, I'd rather give teams the freedom to try something than make them spend weeks proving why they should be allowed to. And if customers don't care? Stop. If it works? Build on it. The cost of learning has changed. Our decision-making processes need to catch up. Source: Scott Kirby, CEO of United Airlines, on Cheeky Pint #Leadership #Experimentation #AI

“I have no idea how to make chips.” That's Elon Musk. And what I find interesting isn't the admission. It's what comes next. He openly admits what he doesn't know. Then he describes how he intends to find out. Study existing factories. Understand the equipment. Start with what's available. Identify bottlenecks. Improve the process. Build custom solutions where they make a difference. That's a very different approach from waiting until you have a perfect plan. I think there's something important here for leaders. Not knowing is not the same as not having a plan. Curiosity to question how things are done. Courage to enter unfamiliar territory. And the discipline to test assumptions, learn from the results, and adjust. None of this guarantees Musk will succeed at making chips. Semiconductor manufacturing is extraordinarily complex. But the approach is worth examining. The job of a leader isn't to have every answer. It's to create the conditions in which the answers can be discovered. And that requires being comfortable saying three things: I don't know. Let's find out. Let's see what we learn. Source: Elon Musk, X Takeover 2026 · Tesla Owners Silicon Valley #Leadership #Innovation #Learning

Nik Storonsky says he's 100% aligned with Revolut. I'm not sure that's even possible. I understand his argument. Founders often have a very different relationship with their companies than professional CEOs. They've built something, taken enormous risks, and invested years of their lives. And yes, professional CEOs can become more concerned with protecting their positions than making difficult decisions. But Nik argues that because he sees himself and Revolut as one entity, his personal interests and the company's interests are identical. A founder can love a company so much that they become the biggest obstacle to its future. Imagine the company needs a different strategy. A different leadership structure. Maybe even a different CEO. Would being 100% identified with the business make those decisions easier? I doubt it. The real test comes when the company needs something different from what its founder personally wants. And sometimes the most important decision a founder can make is to stop being the person who makes every decision. Source: Nik Storonsky, Founder & CEO of Revolut, in conversation with David Senra #Leadership #Founders #DecisionMaking

Abdul El-Sayed just won a Senate seat. Now he has to prove that people were right to trust him. I liked something he said in his victory speech. He thanked the people who knocked on doors, made calls, donated money, and stood by him through difficult moments. And he promised to bring them the kind of leadership that earns their trust. I think that's a good way to look at taking on responsibility. When someone gives you a leadership role, they're making a bet on you. They believe you can do the job. That you'll make good decisions. That you'll be there when things get difficult. But they can't know that yet. I don't think leaders should confuse the confidence people have in them with proof that they've already earned it. And that goes for founders and CEOs just as much as elected officials. Getting the opportunity to lead is one thing. What you do with that opportunity is what people will remember. Source: Abdul El-Sayed, Senator-elect for Michigan, victory speech, November 4, 2026 #Leadership #Trust #Accountability

Ronaldo even thanked his teammates for the arguments. That's not what you usually hear in a farewell message. Cristiano Ronaldo talks about his ambition, the victories, and the defeats that still hurt. But he also thanks the teammates he shared those difficult moments with. I like that. Because I think we sometimes have a strange idea of what a great team looks like. We celebrate trust, collaboration, and team spirit. But when people disagree strongly, we often see it as something that needs to be fixed. I don't. Disagreement isn't necessarily a sign that a team is struggling. The real problem starts when people stop being able to challenge each other, or when an argument becomes more important than the work they're trying to accomplish together. Ronaldo also acknowledges the people behind the scenes who rarely appear in the photographs. And I think that makes his message more meaningful. He's not just looking back at his own achievements. He's recognizing the people and relationships that shaped the journey. You don't need a team where everyone agrees. You need a team that can disagree and still move forward together. Source: Cristiano Ronaldo, official X account, November 5, 2026 #Leadership #Teamwork #HighPerformance #7

Your next rival may redefine your industry. And I think the hardest part is knowing how to respond before anyone knows whether they'll succeed. That's what I find interesting about Tim Höttges' assessment of SpaceX. The CEO of Deutsche Telekom sees the possibility of an entirely new category of communication services, combining satellites, AI, chips, data centers, and mobile networks. He even draws a parallel with Tesla entering the automotive industry. But there's something I appreciate about his assessment. He takes the threat seriously without treating SpaceX's success as inevitable. Building this kind of business would require enormous investment. And established mobile operators still have strengths they can build on. A new competitor can change the rules of an industry without being guaranteed to win. For me, that's where the difficult strategic decisions begin. You have to decide what to protect, what to challenge, and where to invest while the evidence is still incomplete. There is a cost to acting too early. But there is also a cost to waiting. The challenge isn't predicting who will win. It's making sure your organization has options when the rules change. Source: Tim Höttges, CEO of Deutsche Telekom, Q3 2026 Investor Conference Call, November 5, 2026 #Leadership #Strategy #Disruption

Running for head of state?

Why are we asking technology to fix problems that leadership created? Earlier today, I shared my thoughts on Germany's competitiveness. Now here's something from Singapore that I think deserves attention. Prime Minister Lawrence Wong argues that policymakers and technologists need to work together before the important decisions are made. One example says a lot. A government agency had 33 different combinations of penalties for late filings. Instead of developing software to manage them all, they simplified the policy to just four. And I think this goes far beyond government. How often do companies design processes, structures, and rules without involving the people who understand what's technically possible? Then they spend time and money building systems to manage complexity that might never have been necessary. We're using technology to work around decisions we should have questioned in the first place. I'm not suggesting every minister or CEO needs to become an engineer. But if you're responsible for shaping the future of an organization, understanding technology can no longer be someone else's job. Technology shouldn't just execute leadership decisions. It should help shape them. Source: Lawrence Wong, Prime Minister of Singapore, GovTech 10th Anniversary, November 5, 2026 #Leadership #Technology #DigitalTransformation

Germany wants to lead in AI. But we're making it harder for companies to compete. And frankly, I think we're getting in our own way. We have exceptional engineers, industrial expertise, and AI talent. But we also have high energy costs, excessive bureaucracy, and companies questioning whether their next investment should happen in Germany at all. According to the German Chamber of Commerce and Industry's 2026 Energy Transition Barometer, 35% of surveyed industrial companies are postponing investments in their core operations because of high energy costs. Nina Schick, Founder of Tamang Ventures, raises some uncomfortable points about Germany's competitiveness at the Digital Agenda Cyprus Summit 2026. And I agree with much of what she says. But I would go further. We also have a courage problem. Too many leaders are better at explaining why something won't work than taking responsibility for finding out whether it could. Germany urgently needs better conditions for businesses. But we also need leaders willing to make bold decisions, challenge established ways of working, and accept the possibility of failure. You can't regulate your way to innovation. And you can't innovate your way out of every bad regulation. Source: Nina Schick, Founder of Tamang Ventures, Digital Agenda Cyprus Summit 2026 Research: DIHK Energy Transition Barometer 2026 #AI #Germany #Leadership

Imagine spending millions on AI and having no idea what you're getting in return. It sounds absurd. Yet many companies are measuring their AI transformation by how many employees use the tools, how many tokens they consume, or how many use cases they've launched. Alejandro Maza, Chief Product & AI Officer at Kavak, challenges exactly this approach. He distinguishes between three levels of AI usage. At one end, employees use AI tools, but the company has little visibility into the value they create. At the other, AI agents perform specific tasks, and their contribution can be measured against the cost. That's a very different conversation from celebrating adoption rates. But Maza makes another point I find just as important. You need to know what kind of company you're trying to build before deciding where AI belongs. He argues for strong top-down direction. I agree on the direction. Not necessarily on every decision coming from the top. Leaders should be clear about the outcomes they're pursuing. Teams need the freedom to challenge assumptions, experiment, and find better ways to achieve them. Otherwise, we're simply replacing random bottom-up experiments with centrally approved experiments. And neither guarantees a better business. The uncomfortable question for leaders isn't how much AI their organization uses. It's how much of that investment they can actually connect to meaningful results. Source: Alejandro Maza, Chief Product & AI Officer at Kavak, a16z Podcast #AI #Leadership #AITransformation

What happens if people feel they’re no longer needed? Daron Acemoglu, Nobel Prize-winning economist, raises a disturbing possibility. If AI eventually becomes better than humans at almost everything, what happens to our sense of purpose and contribution? That's a hypothetical future. But the decisions that shape our relationship with AI are being made today. And I think leaders need to be careful about the questions they're asking. If every AI initiative starts with “How many people can we replace?”, we're already defining human contribution as a cost to eliminate. What if we started by asking what our people could become capable of with AI? That doesn't mean automation is wrong. Some tasks should absolutely be automated. But there's a difference between removing work that adds little value and designing an organization where people have less and less to contribute. I don't want us to build companies that become more efficient while their people become less capable. Source: Daron Acemoglu, Nobel Prize-winning economist, on 60 Minutes #AI #Leadership #FutureOfWork

Does your best performer get a free pass? One of the most dangerous things a leader can do is make an exception for someone because they're too valuable to lose. Toto Wolff's approach at Mercedes F1 gets to the heart of this. We often measure high performers by what they deliver individually. But what about the people who have to work with them? If someone consistently undermines trust, creates conflict, or makes others less effective, their individual results don't tell the full story. A brilliant person who makes five others less effective might not be your highest performer after all. And this is where leadership has to make a choice. Do you reward the results you can easily measure? Or do you also hold people accountable for the impact they have on everyone around them? For me, high performance has to include both. Source: Toto Wolff, Team Principal & Co-Owner of Mercedes F1, on Cheeky Pint with John Collison #Leadership #TeamCulture #HighPerformance

Nobody wakes up wanting to be a bureaucrat. Yet we keep designing organizations that make bureaucracy almost inevitable. Bill Anderson, CEO of Bayer, shared a revealing example. Over ten years, the number of jobs in an organization he studied stayed roughly the same. But the number of distinct roles doubled. The more narrowly we divide responsibilities, the more dependencies we risk creating. And those dependencies need coordination. We've made managing the consequences of our organizational design a job in itself. Before adding another management layer, I'd want to know how much of that coordination is actually necessary. Source: Bill Anderson, CEO of Bayer, ARC 2026 #Leadership #OrganizationalDesign #Bureaucracy

Are we losing the habit of thinking? I use AI. I see enormous potential in it. But there's a difference between using AI to support our thinking and letting it do the thinking for us. Jim Kwik, brain coach and bestselling author of Limitless, calls the latter “digital deduction.” He points to an MIT study that found differences in brain activity and recall between people writing essays with ChatGPT and those working without AI. The study is small and doesn't prove long-term cognitive decline. But it raises a question worth exploring. Imagine a manager using AI to analyze a difficult decision, assess the risks, and recommend what to do next. The recommendation might be excellent. But can that manager explain the reasoning, challenge the assumptions, and take responsibility for the decision? Because that's what leadership requires, regardless of how good the technology becomes. I want AI to make us better thinkers, not just faster decision-makers. Source: Jim Kwik in conversation with Hala Taha, Young and Profiting, Episode 414 Research: MIT Media Lab, Your Brain on ChatGPT (2025) #AI #Leadership #CriticalThinking