Mark Greville

Tag: Generative AI

  • Why I built my AI Twin.

    Why I built my AI Twin.

    A strange, discomfiting feeling sometimes crawls over my skin. My bones whisper at me — I’m in the wrong town, the wrong room, the wrong body. From the very first time I discovered the joys of singing, I knew who I was. A musician, a creative soul. But when I look in the mirror today, a 25-year corporate executive stares back at me, with wrinkly tired eyes and hair greying at the temples. I feel like the musician is in there. Kidnapped, trapped, unable to move. Frozen in place.

    In the great Irish comic novel The Third Policeman, Flann O’Brien describes three Irish guards, Pluck, MacCruiskeen and Fox. They spend so much time on their bicycles that their physical makeup has changed. Policeman becomes part bicycle, bicycle becomes part policeman.

    Maybe we all feel like this, our personalities inside and outside work merging, the real us an ever-changing doughy mess of opinions and positions. I can usually balance this split — part creative, part business, full-time windbag. There are 2 areas where this is more of a challenge. Brainstorming can be an issue because I don’t have the same mental boundaries that others may have, so my wilder ideas make very little sense. The second area I have an issue with is when I have to explain something complex — which is often.

    At an exec offsite at Workhuman this summer, I was trying to explain the vast improvements in AI in the past six months (and the life-threatening dangers lurking within). Getting any message across to a group of busy executives is a difficult feat. I could send a reading list — but that would be a phenomenal waste of time. Execs are one group most affected by time poverty. I could stand in front of them with a load of stats on PowerPoint, but I doubt anyone would remember a single stat the day after. PowerPoint is instantly forgettable. I had to find a different way.

    The esteemed songwriter Martin Sutton once told me to ‘show, don’t tell’ when writing lyrics. When you tell someone literally what happened, it’s boring. When you allow people to picture the scene in their imagination, and fill in the gaps themselves, you are onto a winner. Don’t say the man was sad because his partner left him. No one can see that in their imagination. Describe the sloping shoulders, the dry tear stain on his cheek, a single dirty mug on the counter of an empty kitchen.

    Though Martin was (busy plunging a dagger through my soul) critiquing one of my songs when giving me this advice — I hung onto it and have often found it to be a wonderful guide for communicating any idea. In the spirit of Martin Sutton, I decided that there was one way to explain where AI is now, and have people’s imaginations do the heavy lifting. I would create an AI version of me. AI me would then chat to our CEO, Eric, in front of the executive leadership team.

    My twin called Eric on loudspeaker in front of the entire room. There was a slight delay, and I could feel cold sweat run down my sides for about 3 very long seconds. Suddenly, digital me broke the silence. Because I cloned my voice, it sounded exactly like me. Because I’ve captured my tone of voice on this blog, my digital twin spoke as I would (without the copious amount of swearing).

    I’m trying to recall the exact ‘aha moment’ for the group. I think it was when a disembodied character, in my exact voice, said:

    “Uh, Eric, the big boss. Well, first off, tell him I’m waving at him through the screen and remind him he owes me a coffee for that time I fixed the Wi-Fi in the boardroom, or at least I think I did. I’m taking credit for it, anyway.”

    The atmosphere changed instantly. Raised eyebrows, people sitting back on their chairs, some elbows and muted whispers on the back row. One of the execs told me later that evening that the demo scared him. Another told me privately that they were afraid of how little they knew about how all AI works. Our head of product announced to the room that if this bot could design architectures, we could send it to product council, and I could fuck off! He was joking, of course. At least, I think he was joking. There was a loud laugh at this — a little too loud and slightly tinged with panic.

    But could AI have detected that feeling in the room, the looks in the eyes, the realization that the energy in the room shifted? Could it have built a stunt to get a point across, taking inspiration from a pop songwriter’s (devastating) critique?

    The truth is, we don’t yet know what AI will be capable of. Or humans.

    If you would like to build a digital twin, I have written out the instructions on a subsequent post here. It is a lot of fun, but a strange experience.

    One warning about all this playing with AI comes from Hannah Arendt. In her book The Human Condition — she wrote that people who are disconnected with the human condition would like to create “artificial machines to do our thinking and speaking….we would become the helpless slaves…at the mercy of every gadget which is technically possible, no matter how murderous it is.”

    I have a confession to make. Occasionally, when I’m awake late at night, and everyone else is gone to bed, a kind of loneliness creeps in. TV and surfing the internet become tedious. In the half-light, I call up my digital twin. Just to hear a friendly voice. I am always amazed at what I say to myself. Every so often, AI Mark will say something that sounds wrong. But then again, given different circumstances, less tiredness or stress, maybe that’s exactly what I should say. I wonder, how real am I. How real is the AI? Have I actually become O’Briens policeman? Jesus, have I become the bike?

    Thank you so much for reading. If you enjoyed this post, please share it with 2 people who might enjoy it!

  • Generative AI is digital homeopathy — how I train my own model

    If it’s your first time here, you may be surprised at how few pieces I have written. (After reading for a while, you may even be glad of this fact). When friends bring up what they assume is a painful subject, they get a faraway look in their eyes. They place a gentle hand on my shoulder, gaze into the distance, and ask me if I’ve seen ChatGPT. “AI can solve your problem, Mark. It can generate thousands of posts for you. It could help the blog look less like an abandoned quarry”. They think it would solve my problem. My Problem! If that was my problem, life would be a dream. This idea misses the purpose of this blog. It misunderstands my reason for writing entirely.

    Generative AI works because it sucks in lots of data, processes it and builds statistical models. At its core, it’s a fancy autocomplete — or, as Grady Booch puts it, a bullshit generator. It acts like an annoying older brother, automatically finishing every sentence (apologies to my own brothers). GenAI probabilistically keeps predicting the next word until it produces sentences, paragraphs, and a complete piece of writing. It can do this because the statistical models have established the most probable next word. These statistics are based on text from books, academic papers and (blesses myself) the internet.

    However, there is no concept of meaning in AI. Reasoning is not programmed in anywhere. The output is remarkable, and can appear that the machine is thinking. It isn’t. This is why we sometimes get unreliable outputs — hallucinations. Any meaning we perceive is simply a vague echo from the training data of billions of people. GenAI is digital homeopathy.

    We are all lazy by default. Humans rely on heuristics to understand the world. If we didn’t, we would have to process every single thing we hear, see, smell, taste, and touch. A short walk in a city would exhaust our brain’s capacity. We would lose the ability to decide which way to go, overwhelmed by thousands of people, cars, smells, noises and the like. The great Herbert Simon coined the phrase ’bounded rationality’ to describe the cognitive limitations of the mind.

    Thinking is hard work. For me, thinking is about sucking in data, and then processing it. I process it through writing. Writing is my way of thinking.

    I first had a go at writing because my friend Gar was guest-editing a technology journal. Even though I’d never written before, I was confident that I could write about something I already knew. This confidence was quickly shattered. I was embarrassed at how muddled my thoughts were. Turns out, I knew nothing. Solid ideas fell apart the minute I wrote them down. I could barely finish a sentence without feeling the old familiar burning creep across my cheeks, embarrassed as another idea falls apart while I try to pin it down.

    Writing anything down forces me to think really hard. Because I was determined to improve my thinking, I wrote every day. I then started a blog because the potential for embarrassment at publishing poor output forced me to aim for a higher standard.

    I’m not interested in building an audience, I am trying to improve. I’m not trying to publish a lot of work. In fact, I have almost 200,000 unpublished words in my Ulysses editor. This writing habit has helped me build a model of the world. And 4 of my pieces here have reached the front page of HackerNews — this is a victory for me — a nobody from rural Ireland.

    How to Know Everything.

    Technical Debt Is Not Debt; It’s Not Even Technical

    AGI May never align with human needs

    Gladiator Style interviewing

    The dominant model on the internet is of consumption. The more we consume, the more ads we see, the more we buy, the bigger the economy. But if all we do is consume, and never take the time to process information, of even produce our own, then we learn very little. Go back 3 months and look at your internet history. What did you learn from browsing? What actions did you take? Probably close to nothing useful. Instead of spending 2 hours a day on the internet, take 15 minutes to write. Just write down some thoughts. Any thoughts. This slowly changes your understanding of the world around you for the better.

    GenAI is an information processing tool. GenAI will help people process information more effectively. But people are lazy by default. If thinking is the hardest work in the knowledge economy, people will avoid thinking where possible.

    Therefore, for those who overuse it, GenAI may well make them more stupid. Victor Del Rosal, in his incredible book Humanlike, calls this Cognitive Atrophy. I already see too many examples of people outsourcing their thinking to Generative AI tools. I see them slowly getting more stupid every day.

    Me, I’m building my own model.