Midnight Musings

Author: thebaconbaba

  • Stop Asking What AI Can Do

    Stop Asking What AI Can Do

    Most people ask the wrong question about AI.

    “What can AI do for me?”

    That’s like hiring someone and then asking what skills they happen to have. Nobody competent works that way.

    You start with the job. You define what you need. You break the work down. Then you hire for that. And once you hire, you train. You give context. You correct. You refine.

    AI is no different.

    Look at your actual day. Status updates. Follow-ups. Notes. Deck edits. Repeating yourself in different formats. Break that chaos into the smallest possible tasks and hand them off.

    AI doesn’t need inspiration. It needs instructions. And a bit of training. The clearer you are about how you work, the more useful it becomes.

    Used this way, it’s not a replacement. It’s leverage. It doesn’t change who you are. It just gives you more surface area to operate.

    Stop admiring the tool. Start assigning it work.

    #midnightmusings from the trenches of delivery.

  • Most Teams Don’t Have a Technology Problem

    Most Teams Don’t Have a Technology Problem

    One of the strangest habits in delivery is how quickly troubled programs blame technology.

    Deadlines slip?
    Must be the platform.

    Escalations rise?
    Probably architecture.

    Delivery slows down?
    Clearly a tooling issue.

    But most program failures are not caused by technology limitations.

    They come from unclear ownership, delayed decisions, competing priorities, and teams that slowly normalize confusion.

    The response is usually predictable:

    • another tracker
    • another status call
    • another governance layer

    Because process feels safer than accountability.

    Technology becomes the visible villain because it is easier to debug systems than confront operating behavior.

    Most struggling programs already have good enough technology.

    What they lack is operational clarity.

    Clear ownership.
    Faster decisions.
    Less ambiguity.

    Simple.
    Difficult.
    Rare.

    #Midnightmusings from the trenches of delivery.

  • Right Person. Wrong Role.

    Right Person. Wrong Role.

    One of the hardest parts of leadership is accepting that good people can still be wrong for a role.

    I once heard Girish say:
    “Right person for the right job.”

    Simple sentence. Difficult responsibility.

    Because eventually every leader faces the same uncomfortable reality:
    the person may be hardworking, loyal, and trying their best – and still not be the right fit anymore.

    You see it slowly.
    Missed ownership.
    Repeated escalations.
    The team quietly compensating in the background.

    And this is where leaders hesitate.

    Not because they don’t see the problem.
    Because they do.

    They delay the conversation hoping time will solve what clarity already knows.

    But keeping the wrong person in the wrong role too long is unfair to everyone involved – especially them.

    Hard decisions do not require emotionless leadership.
    They require calm leadership.

    Be prepared with data.
    Be clear.
    Don’t over-explain yourself.

    The best leaders handle difficult decisions quietly.
    No drama.
    No corporate theater.
    Just clarity.

    Because delayed decisions rarely become easier.
    They usually become expensive.

    #midnightmusings from the trenches of delivery.

  • When the Vatican Starts Writing About AI

    When the Vatican Starts Writing About AI

    Pope Leo XIV just released a 245-page encyclical on Artificial Intelligence.

    You know things are getting serious when a 2,000-year-old institution decides AI needs formal doctrine.

    The Vatican’s new AI encyclical is not really about technology. It’s about power, labor, truth, identity, and what happens when human intelligence itself becomes industrialized.

    That’s the shift.

    AI is no longer a “tech trend.”
    It is becoming infrastructure for society itself.

    The Church has historically stepped into moments where technology reshaped humanity:

    • Industrialization
    • Nuclear weapons
    • Global capitalism

    Now AI joins that list.

    And beneath the religious framing sits an uncomfortable secular reality:
    Every major institution now understands AI is going to fundamentally alter how civilization operates.

    Governments.
    Education.
    Media.
    Law.
    Work.
    Trust.
    Human agency itself.

    This is no longer a conversation about productivity tools or chatbot demos.

    Once institutions built to think in centuries start treating AI as a moral and societal question, you are no longer in an innovation cycle.

    You are in an epoch shift.

    Humanity, naturally, appears determined to navigate this transition with deep wisdom and maturity. Right after deepfake propaganda, autonomous weapons, and emotionally dependent chatbot relationships. Spectacular species behavior.

    #Midnightmusings from the trenches of delivery.

    Visuals by AI. Reflections by experience.

  • AI Is Creating a New Class of Program Managers

    AI Is Creating a New Class of Program Managers

    For years, program management optimized around coordination.

    Status calls.
    Follow-ups.
    Approvals.
    Escalations.
    Dependency tracking.

    A large part of program management became operational middleware between disconnected teams, systems, and stakeholders.

    AI is starting to change that.

    Not because it replaces delivery managers.
    Because it commoditizes execution support.

    Presentations, summaries, reporting, analysis, documentation, planning drafts. Machines can now generate acceptable first versions in seconds.

    The advantage is shifting elsewhere.

    Toward delivery leaders who can:

    • define problems clearly
    • reduce ambiguity
    • make tradeoff decisions
    • simplify complexity
    • align execution across teams

    AI rewards clarity.
    Not activity.

    The highest-value delivery managers are no longer the people producing the most artifacts.

    They are the ones creating the most alignment.

    Because one clear operator with AI can now drive execution with leverage that previously required layers of coordination, meetings, and process overhead.

    The role is evolving faster than most organizations realize.

    And the shift has already started.

    #Midnightmusings from the trenches of delivery.

  • Being Human

    Being Human

    Just got back home from a Leadership AI Summit at NICE.

    Leaders across Product, R&D, Services, and Support spoke about how the workforce is evolving and how processes are evolving to adapt to an AI-driven world.

    There was a lot of healthy discussion around AI-enabled delivery, AI-DLC, workforce transformation, domain breadth, and solution expertise.

    But one thing stood out clearly through all of it.

    While technology evolves, one thing still stays the same.

    The customer experience.
    The human touch.
    Empathy.
    Communication.
    Soft skills.
    The ability to build trust, calm uncertainty, and connect with people.

    That remains irreplaceable.

    AI can accelerate execution.
    It can summarize, automate, recommend, and optimize.

    But it still cannot truly replace the human ability to understand context, navigate emotion, and build relationships during moments that matter.

    Well… not yet at least.

    Ironically, the more AI advances, the more valuable these deeply human skills become.

    That may very well be the new gold.

    #Midnightmusings from the trenches of delivery.

  • AI Has Changed The Cost of Waiting

    AI Has Changed The Cost of Waiting

    In the last post, I wrote about how the people winning with AI aren’t necessarily the best coders.

    They’re the people who understand their domain deeply enough to build.

    But there’s another shift happening underneath that.

    Speed.

    A few months ago, I watched two very different approaches to the same AI-driven idea.

    One treated it like a traditional software project:
    planning, reviews, alignment, architecture discussions, phased execution.

    The other approach was simpler:
    build fast, get it into people’s hands, refine as you go.

    That contrast stayed with me.

    Because AI is collapsing the distance between idea and execution.

    A domain expert with clarity and the right tools can now prototype faster than many organizations can align internally.

    And that changes things dramatically.

    The advantage is no longer just technical skill.

    It’s speed of understanding.
    Speed of iteration.
    Speed of decision-making.

    Many organizations are still operating with waterfall thinking in a world where experimentation has become almost free.

    AI rewards people who are hands-on.
    People close to the actual business problem.
    People willing to fail fast and refine in public.

    Which raises an uncomfortable question:

    If everyone starts building this quickly, what happens to stability, architecture, governance, and long-term maintainability?

    That’s probably where the real conversation begins.

    #Midnightmusings from the trenches of delivery.

  • The People Winning With AI Aren’t the Best Coders

    The People Winning With AI Aren’t the Best Coders

    That’s the wrong conversation.

    A few years ago, building software meant technical skill. You needed engineers, architects, specialists, infrastructure teams, databases, deployment pipelines – the whole machinery.

    Now?

    You can build surprisingly complex workflows, dashboards, automations, apps, even lightweight platforms using plain language. Tools like ChatGPT, Gemini, Claude, Cursor, Lovable, Replit – they’ve flattened the technical barrier faster than most organizations realize.

    The bottleneck is no longer execution.

    It’s understanding.

    A few weeks ago, Girish made an observation that stuck with me:

    “Knowing your domain is becoming more important than being technically excellent.”

    And honestly, that might be the biggest shift AI is creating right now.

    The people getting ahead with AI aren’t necessarily the best programmers. They’re the people who deeply understand their domain. Their customers. Their workflows. Their operational pain points. Their industry logic.

    Because AI can generate code.

    But it cannot invent clarity.

    If you truly understand how your business works, you can now describe it, structure it, refine it, and have AI build around it at absurd speed.

    That changes the game completely.

    The value is shifting from “Who can build?” to “Who can think clearly enough to design what should be built?”

    Ideas are becoming leverage.

    Context is becoming leverage.

    Conceptualization is becoming leverage.

    Execution is slowly turning into the cheaper commodity.

    And that creates a second shift that most organizations still haven’t fully understood:

    When execution becomes easier, speed starts mattering more than process.

    The teams that learn fastest may soon outperform the teams that plan the longest.

    And honestly, that changes leadership, delivery, and product development more than AI itself.

    More on that in the next post.

    #Midnightmusings from the trenches of delivery.

  • Ruthless Simplicity. Relentless Execution.

    Ruthless Simplicity. Relentless Execution.

    Most delivery problems don’t start with technology. They start with process drift.

    Over time, organizations quietly accumulate layers – another tracker, another template, another governance step added after a crisis. None of it feels unreasonable in the moment. But eventually delivery teams spend more time navigating process than delivering outcomes.

    At some point, someone has to ask: Is this actually helping?

    Recently, our COO Arun Chandra framed operational excellence around three principles – Ruthless Simplicity, Crystal Clear Accountability, and Relentless Execution.

    Simple words. Hard in practice.

    Because Simplicity forces you to remove things. Accountability forces you to name owners. Execution forces you to stop admiring frameworks and start delivering.

    As part of our #AIFirst initiative, we redesigned the NICE Actimize XSE delivery governance model.

    Instead of adding reporting layers, we introduced an AI-driven governance layer across the delivery lifecycle – analyzing signals from risks, timelines, and project updates to surface issues early.

    In practice, it meant collapsing multiple trackers into a single lifecycle model and letting AI highlight emerging risks before they become escalations.

    The goal isn’t more governance.

    It’s better visibility with less friction.

    #Midnightmusings from the trenches of delivery.

  • The Smartest Tool in the Room

    The Smartest Tool in the Room

    This thought began while we were evaluating new Professional Services Automation (PSA) platforms — every vendor pitching their “AI-powered, all-in-one” solution to simplify delivery, optimize resources, and predict success before kickoff.

    Sounds brilliant. Except every program manager knows the truth: none of these tools truly talk to each other.

    We live surrounded by “smart” systems — Asana, Changepoint, Smartsheet, Jira, Salesforce, Monday.com, Google Workspace — yet we still spend hours stitching them together. Each tool works in isolation, but together? They’re chaos wrapped in APIs.

    That’s why, even in 2026, Excel remains the command center. It’s where all the scattered data finally makes sense. Because no matter how advanced the tech gets, AI still can’t replace human judgment, context, and the ability to simplify.

    The goal isn’t AI everywhere — it’s clarity everywhere.

    Until then, Excel and human judgment remain the most reliable AI we’ve got.

    — Midnight musings from the trenches of delivery.

    #Midnight-Musings