michael-dean-k/

We now have products that scan family members to turn them into posthumous avatars. The tagline: “With 2wai, three minutes can last forever.” It's weird to have this so soon. As someone who is down with a posthumous digital consciousness that my kids can interact with, I even find this to be too weird for me. The problem that it uses video to serve as a replacement for a deceased relative. A few boundaries that are important for me:

1. By keeping it text-based instead of video, it’s more like you’re interacting with a proxy of my mind instead of my body/soul. It won’t register in my child’s brain as “me” and so it will be less confusing, less toxic to the grieving process. 
2. It should refer to me in the third-person, even if it is trained on me and sounds like me. It should not be an imposter of me, but a proxy/guide of my thoughts/beliefs, almost like an elder guide.
3. It should cite my original logs/essays/journals. In effect this makes the experience similar to something we already have: reading your grandparents journals. This just makes it possible for your questions to immediate summon the relevant wisdom.

The comment section was in unanimous agreement:

• This is one of the most vile things I’ve seen in my life.
• You are a psychopath.
• Shoot that guy.
• You’re creating dependent and lobotomized adults by doing this.
• Demonic, dishonest, and dehumanizing.
• Hey so what if we just don’t do subscription-model necromancy.
• Oh goody, another way for people to completely lose touch with reality and avoid the normal process of grief.
• Nightmare fuel.
• I don’t see how people can say demons aren’t real when there are beings around us willing to create shit like this.
• “You will live to see manmade horrors beyond your comprehension.” — Tesla.

I’d say this is an extremely lightweight microcosm of the core dilemma of what the 2040s will face: a moral war over technology that changes the constraints of human life.

1. If you can suspend disbelief, this AI ad is a good example of using video to normalize a concept
2. I’m skeptical of the promise here, not because digital twins will eventually become a thing, but because I doubt you can create a high-resolution twin in 20 minutes. Whether you write or speak to it, how many words will be generated, 1,000? The matches will be surface level, maybe slightly more than what’s already on an active LinkedInIn account. For this to be useful, you need more like 100x the data (someone like OpenAI would be more likely to pull this off).

“In just 20 minutes, your AI gets to know you: your goals, your talents, your quirks, your questions. We build a private, structured map of who you are—and what you’re seeking. This is your Twyn. It’s like a digital twin, but proactive. Every day, your Twyn holds thousands of intelligent conversations with other Twyns in our global network. It explores who they are, what they offer, and what they need—and looks for meaningful overlap with you. When something clicks, you hear about it. _Not spam. Not noise. Just signal. A founder meets their first investor. A coach finds a client they can truly help. A traveler finds a local guide who shares their values. A lonely genius finds someone who finally gets them. This isn’t networking. This is serendipity—on demand.”

I imagine there will be two kinds of agent companies: 1) for modular services (ie: generic SaaS coderbots for $1k/month), but also, 2) for replicating identity. In a world where everyone gets access to swarms of modular agents that move at unimaginable speeds, I think those with a “replica” will have the edge. A replica is like a centralized manager entity—one that has your corpus of principles, lessons, memories, ideas, taste, libraries, etc.)—to coordinate between all the agents.

If this comes true, it brings new leverage to writing out your thoughts in detail. Maybe for every 10k words I write, my agent swarm is 2x more aligned to my specific goals. However, it also inverts the norms of data privacy. There’s a trade-off: the more you use writing for distribution and audience-building, the more clonable you become. It’s causing me to think harder about when a piece of content is public/gated/private.

So if models, apps, and infrastructure become too cheap to matter, then the only thing left that’s scarce is personal data / personal perspective. You’d think Notion/Google are positioned to solve this, but it’s interesting to watch people use Cursor as their PKM system because it’s so easy to connect personal context with an array of models.

David and I got into this talk on how technology will bring us to a point of serious disorientation, and how a Christian (or religious/moral) framework can help ground all the turbulence. At some point I made the point that I think it’s important to die. I believe that some kind of soul-flight is the main climax of consciousness. I don’t believe in the Christian eternal afterlife, but in a 3-minute moment of time dilation that is, in essence, an eternity—1,000 lifetimes. As so while I never want to be literally immortal, I want literary immorality. All the text of my life establishes a moral foundation for a text-based replica to outlive me and interact with and hopefully guide future generations, whether it’s 10 or 10 million people. Books already fill this function, except now books are agentic and interactive.

Feels like multiple AI writing apps nonchalantly launched a "persona cloning" feature. Just upload a Substack archive URL and you can mimic your voice! (Or, someone else’s?) For now, AI mimicry is pretty hollow for anything serious. But what about when it’s not? I think we’re heading towards a paranoia where more people paywall their writing to prevent scraping.