Productivity

How much physical energy does it ACTUALLY cost you to worry about the future?

Gartner just surveyed 350 large enterprises deploying AI. 80% cut jobs. Some by as much as 20%. The result? The companies that cut the most showed nearly identical financial returns to the ones that cut the least. In several cases, the ones that cut less performed better. No correlation between AI-driven layoffs and improved ROI. None. Gartner's Helen Poitevin was direct: "Workforce reductions may create budget room, but they do not create return." Cutting people frees up cash. It does not generate value. Most leadership teams are conflating the two. So what actually works? Upskilling staff to work alongside AI. Redesigning roles around what humans do well vs. what AI does well. Building operating models where people guide autonomous systems instead of getting replaced by them. There's a real difference between using AI to do the same work with fewer people and using AI to unlock work that was previously impossible. The first saves money on paper. The second compounds over time. We've already seen the pattern. Klarna cut 700 CS roles, watched quality decline, and started rehiring. IBM automated HR functions and reversed course. The Commonwealth Bank of Australia reversed 45 AI-driven layoffs after realizing those roles were never redundant. Gartner predicts half of companies that attributed headcount cuts to AI will rehire under new titles by 2027. If someone in your org is building an AI business case around headcount reduction, share this data. The assumption that fewer people equals better margins equals better returns is not supported by the evidence. AI is not leading to a jobs apocalypse. It's changing the shape of what people do. The companies that understand that difference will be the ones worth working for, and buying from, three years from now. Read the full piece on State of Brand here: https://lnkd.in/ggH-NXyM

Motivation is the least reliable driver of behaviour change. It fluctuates. A good morning, a bad week, a stressful month, motivation moves with all of it. And most people experience this. They wanted something, started, fell off, and concluded they didn't want it enough. Wendy Wood, a psychologist at USC, studied this for 30 years. She found that the people who changed most durably were the ones who changed their environment until the behaviour became automatic. It had nothing to do with how motivated they were. The gym on the commute home. The kitchen with healthy food in it. Small contextual shifts that made the right behaviour easier to do than avoid. 43% of what people do every day isn't a decision. It's a context firing. Same place, same time, same cues, the brain automatically starts executing. Once a behaviour crosses that threshold, motivation becomes largely irrelevant. The context carries it. The people who relied on motivation alone kept cycling. High intention, inconsistent follow-through, same conclusion every time. The ones who lasted stopped trying to sustain motivation and started building situations where they won't need to rely on it. Motivation will get you started, but your environment will decide what stays. #rajshamani #figuringout

Every person should have one folder in their computer that is the "In Case of Death" folder... and this is how it is structured... The folder is used to manage and pass on one's assets to successors easily... and to avoid any legal hassles. This folder is divided into 5 parts: 1/ Personal IDs - Birth Certificate - Marriage Certificate - Aadhaar and Passport - Divorce Decrees - Change of Name Affidavits 2/ Legal & Estate Management Documents - Last Registered Will with notarized copy - Succession Certificates for Movable Assets - Guardianship Documents for minor children / dependants 3/ Financial Assets and Accounts - Bank Account details and statements - Demat Account, Shares, Mutual Funds, Investment Portfolio - Insurance Policy Documents with Beneficiaries - Outstanding loan documents - Bank Locker or Safe Deposit Locations and Access details 4/ Property Documentation - Real Estate and Business - Property Deeds and Titles for Real Estate - Mortgage Documents - Rental Agreements - Business Ownership Documents and Succession Plan 5/ Digital Assets and Access Information - Email with recovery information - Social media contact and legacy contact preferences - Domain names, Cloud Storage - Digital Subscriptions of material value The best way to manage passwords here is to provide access to Password Managers where you can add an Emergency Contact who can access all passwords in case of such conditions. There are apps like 1Password, NordPass and Keeper which offer printable PDF documents for emergency access that can be kept in this folder. These apps allow a waiting period from last activity after which the emergency contacts can use the data. Also, please ensure that you submit all such documents for probate and get them drafted by a trusted lawyer. I would also hope you would tell trusted family members about where you have stored this information so that they can access it when required. Above all, wish you all a long and healthy life. #casarthakahuja

“Stop calling everything an agent” Folks are slapping 'agent' on anything that touches AI. It’s confusing. I jumped on Peter Yang’s podcast to unpack why that’s misguided, and why the middle ground (agentic workflows) is where real ROI lives. Workflows = determinism and reliability Agents = judgment and flexibility Stitch them together and you get the safest, most powerful form of AI orchestration. Give your agent just enough tools and context to do one job exceptionally well, and orchestrate the rest with workflows. Two examples of this I’m using personally: 𝗘𝘅𝗲𝗰𝗖𝗼𝗻𝗻𝗲𝗰𝘁 (Workflow / Determinism)  👉 A teammate submits an Interfaces form to request exec engagement, the AI drafts an on-brand message, posts to Slack, and tracks it in Tables. This one runs hundreds of times a month. Zero chaos, total clarity 𝗜𝗻𝗯𝗼𝘅 𝗠𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁 (Agent / Inference) 👉 When a new email arrives, the agent reads it, checks HubSpot, uses reasoning to categorize it, and sorts it in Slack. It flags customers, intelligently routes the right things to my EA Cortney, and clears all of the noise. It reasons like a human assistant, saving us hours each week with full transparency logs Cortney’s Inbox Management agent is actually available as a template, you can steal it here: https://lnkd.in/gR8UjTCN Full episode coming soon. Thanks for having me, Peter.

Readers responded with both surprise and agreement last week when I wrote that the single biggest predictor of how rapidly a team makes progress building an AI agent lay in their ability to drive a disciplined process for evals (measuring the system’s performance) and error analysis (identifying the causes of errors). It’s tempting to shortcut these processes and to quickly attempt fixes to mistakes rather than slowing down to identify the root causes. But evals and error analysis can lead to much faster progress. In this first of a two-part letter, I’ll share some best practices for finding and addressing issues in agentic systems. Even though error analysis has long been an important part of building supervised learning systems, it is still underappreciated compared to, say, using the latest and buzziest tools. Identifying the root causes of particular kinds of errors might seem “boring,” but it pays off! If you are not yet persuaded that error analysis is important, permit me to point out:  - To master a composition on a musical instrument, you don’t only play the same piece from start to end. Instead, you identify where you’re stumbling and practice those parts more. - To be healthy, you don’t just build your diet around the latest nutrition fads. You also ask your doctor about your bloodwork to see if anything is amiss. (I did this last month and am happy to report I’m in good health! 😃) - To improve your sports team’s performance, you don’t just practice trick shots. Instead, you review game films to spot gaps and then address them. To improve your agentic AI system, don’t just stack up the latest buzzy techniques that just went viral on social media (though I find it fun to experiment with buzzy AI techniques as much as the next person!). Instead, use error analysis to figure out where it’s falling short, and focus on that. Before analyzing errors, we first have to decide what is an error. So the first step is to put in evals. I’ll focus on that for the remainder of this letter and discuss error analysis next week. If you are using supervised learning to train a binary classifier, the number of ways the algorithm could make a mistake is limited. It could output 0 instead of 1, or vice versa. There is also a handful of standard metrics like accuracy, precision, recall, F1, ROC, etc. that apply to many problems. So as long as you know the test distribution, evals are relatively straightforward, and much of the work of error analysis lies in identifying what types of input an algorithm fails on, which also leads to data-centric AI techniques for acquiring more data to augment the algorithm in areas where it’s weak. With generative AI, a lot of intuitions from evals and error analysis of supervised learning carry over — history doesn’t repeat itself, but it rhymes. [Truncated due to length limit. Full text: https://lnkd.in/gjqv6VeA ]

Multi-Head Attention (MHA) is the engine of LLMs. But over the years, we have added several tweaks to make it more efficient for long-context settings, especially when using KV caching during inference. I implemented the most common variants from scratch: 1) Grouped-Query Attention (GQA): Instead of having a unique key and value for each query head, multiple queries share the same key and value. As long as the sharing ratio is not too extreme, this has minimal impact on model quality. It is the most widely used variant today and found in almost every modern LLM including Llama 2-4, GPT-OSS, Gemma 3, Qwen 3, GLM 4.6, and many others. 2) Multi-Head Latent Attention (MLA): This variant introduces a compressed latent representation for the keys and values that are stored in the KV cache. During inference, these latent keys and values are up-projected back to the full dimension. The extra projection adds a small computational cost, but the memory savings make it worth it. This approach is currently used by DeepSeek V3 and Kimi K2. 3) Sliding-Window Attention (SWA): SWA restricts each token’s attention span to a fixed local window, which reduces memory needs by shrinking the KV cache in long-context regimes. It is usually applied selectively, for example every other layer, or in Gemma 3's case, five SWA layers for each full-attention layer. While less common today, it remains an important optimization, notably in Gemma 3. All three variants, GQA, MLA, and SWA, can also be combined freely within the same model. Here's a link to check them out: 1️⃣ GQA: https://lnkd.in/grDPXUUi 2️⃣ MLA: https://lnkd.in/gm4FzE32 3️⃣ SWA: https://lnkd.in/g7x-fdgn

Most people are taught how to be high performers. But too few are taught how to perform in a team. And that’s a problem, because in most roles, you’re not an individual contributor. You’re part of a larger entity, working with others to build something. Yet, I see founders spend hours refining their product or systems,  But don't devote time to team development. At HomeServe, I approached team performance with purpose,  And it was one of the best decisions I made. Here are 7 tools I’ve used (and still use) to build high-performing teams,  Based on real lessons from building a £4.1bn business: 1️⃣ Start With Why (Simon Sinek) ↳ Before you focus on what or how...get clear on why. WHAT – The product you sell or the service you provide HOW – What makes you different WHY – Your deeper purpose or belief Every great team needs a reason to get out of bed in the morning. 2️⃣ The 70-20-10 Rule (McCall, Lombardo & Eichinger) ↳ How people actually learn on the job: 70% from challenging experiences 20% from coaching and mentoring 10% from formal training Most teams over-invest in training, and under-invest in real development. I'm amazed at how few founders or CEOs have a coach or mentor. 3️⃣ The Trust Triangle (Frances Frei, Harvard) ↳ Trust isn’t built with perks. It’s earned in three ways: Authenticity – Are you real? Logic – Do your decisions make sense? Empathy – Do you care? Without trust, you can’t build speed or loyalty. 4️⃣ The 5 Stages of Team Development (Tuckman Model) 1. Forming – Team gets together 2. Storming – Conflicts surface 3. Norming – Ground rules form 4. Performing – Results roll in 5. Adjourning – Project ends or evolves Don't panic during ‘storming’. It’s necessary friction. 5️⃣ The Johari Window (Luft & Ingham) ↳ Self-awareness is a team sport. Open – You know, they know Hidden – You know, they don’t Blind Spot – They know, you don’t Unknown – No one knows (yet) This helps surface feedback, build confidence, and avoid surprises. 6️⃣ The Energy/Impact Matrix (Inspired by McKinsey) ↳ Map every team member’s impact vs. energy. Use it to: Make smart hiring/firing decisions Spot burnout early Retain high performers High-performing teams don’t tolerate drift. 7️⃣ The RAPID Decision-Making Model (Bain & Company) ↳ High-performing teams make fast, clear decisions. Recommend – Suggest the course of action Agree – Those who must sign off Perform – Executes the decision Input – Provides relevant facts or opinions Decide – Final decision-maker This clears up delays, dropped balls, and blame. Building a great team is about building an environment where talent can actually thrive. I go deeper into team-building in my new book. Order it today: https://lnkd.in/eRYDKXdT ♻️ Repost if you believe team performance should be built, not assumed. And for more on how I scaled teams to build a £4.1bn business, Follow me Richard Harpin.

Rick Rubin went on stage in Helsinki the day after my talk. Someone asked how he resolves creative differences with artists. His answer was simple: change the conversation from "I disagree" to "let's build it." Then he shared a story: An artist played him a song. The transition didn't work. Rubin told him so. The artist said, "We'll just cut that part in half." Rubin thought to himself: What a dumb idea. But he didn't say that. He said, "Let's try it." The artist played it. It worked. Rubin is a legend. He's produced everyone from Johnny Cash to Jay-Z. Instead, he bit his tongue and let the artist prove him wrong. The principle: when you make an idea tangible, it stops being the person's idea. It becomes something you can both look at objectively and improve together. Once you build it, the truth is obvious. Here's what this looks like in practice: Your designer wants to change the entire homepage layout. You think it's too risky. Instead of three meetings debating it, you say: "Let's build a prototype and test it with 50 users this week." Your sales team wants to restructure the pricing page. Instead of blocking it because you're worried about conversions, you say: "Let's run it as an A/B test on 20% of traffic for two weeks." Your engineer wants to rebuild a core feature from scratch. You think it's overengineered. But instead of killing it in the planning phase, you say: "Spike it out in three days and show me if the performance gain is real." You're not saying yes to everything. You're saying, "Let's find out." Rubin also said something that stuck with me: "If there's disagreement, I always side with the artist's vision. Because to them, it's their career. To me, it's just one piece of my portfolio." Most leaders think backing down makes them look weak. Rubin knows that siding with the person who has the most at stake makes better work happen. Your job isn't to be right. It's to create the conditions where the best idea wins. Stop debating. Start building. P.S. This insight is from this week's newsletter where I break down why Yamaha dominates while Steinway got sold to private equity: https://lnkd.in/efSqP_9K P.P.S. Access additional research links, the podcast, and the full archive in the first comment 👇 Thank you to Nordic Business Forum!

⚡️ What if the future of IT wasn’t about resolving tickets… but eliminating them altogether? 🎫 In IT, the ticket has long been the center of the universe. Every issue. Every request. Every alert. It all starts with “open a ticket.” But here’s the problem: humans create tickets… and now machines do too. With endless monitoring, alerts, and logs, the machines are literally yelling at us and flooding IT teams with noise. 🔥 In my conversation with Sean Heuer, CEO of Resolve, he explains how AI + automation + orchestration can flip this model upside down: 1️⃣ Own the front door → intercept tickets before they’re created. 2️⃣ Aggregate the noise with AIOps into one clear stream. 3️⃣ Apply AI to reduce volume and automation to self-remediate most issues. This is the path to zero-ticket IT—a world where the focus shifts from fixing tickets to delivering better employee experience and healthier systems. 🎥 This clip is just a taste. Watch the full interview here: https://lnkd.in/e2ZcizJJ Curious about agent? https://lnkd.in/eaRMtsAe #AI #Automation #FutureOfWork #ZeroTicketIT #DigitalTransformation #ITOperations #AIAgent