programming

Incorruptible Organizations AMA with Eric Ries. Wednesday 2/4 at 3:00 PM CT

Lean Startup author who now focuses on legal structures to protect mission-driven organizations from corruption. incorruptible.co

Free book giveaway! Register here.

AMA with Jeffrey Ladish. Wednesday 2/4 at 2:00 PM CT

Executive director of Palisade Research; studying AI loss of control risks.

AMA with Jeffrey Ladish. Wednesday 2/4 at 2:00 PM CT

Executive director of Palisade Research; studying AI loss of control risks.

Hey Gang. Hey there! I am a children's book author, part time travel agent and sports nut. Hope everybody's hanging in there. 

Please help me stay intellectually honest! I'm not a fan of generative AI in general, and LLM technology specifically. I think its capabilities are being drastically over-hyped. It's a perfect, sweaty example of a solution looking for a problem. I'm skeptical of many claims people are making wrt how it's helping them.

My experience is it's like having access to an idiot-savant intern. Awful at most tasks, but knows everything and can read incredibly quickly.

Publicly, I've taken on the mantle of a staunch critic of generative AI and a pro-human, pro-soul advocate.

And for the most part, I'm happy with that stance. I like it. It feels good to rail against something, and it feels good to contrast a thing that I hate against something I love. It throws the love into more relief.

Yet, I don't want to lose any babies in that bathwater, and I don't want to lose my intellectual honesty in the neurochemical rush of fighting for a cause. So I'd love to explore the best use cases of LLMs that you all are actually using, and actually finding beneficial, life improving, productivity increasing, all of that.

I'd love to hear your experience, and ideally, you'd to tell me how you're doing what you're doing with it in enough detail so that I can try it.

I'll start.

Absolutely most useful thing I've found for it so far, and it's not even close, is language learning.

I'm in a slow process of learning Japanese, and asking a chatbot to break down the grammar of a specific sentence is super useful. It's also great for generating content for flashcards. Say you have a set of characters, and you want some example words that use each particular character. It's so easy to generate stuff like that.

Outside of that, I use it in super basic ways (basically as google with one less step).

So please, give me your best use cases, things that you've not only been impressed by, in a "oh wow, that monkey can tap dance!" way, but that has actually improved the quality of your life.

Please help me stay intellectually honest! I'm not a fan of generative AI in general, and LLM technology specifically. I think its capabilities are being drastically over-hyped. It's a perfect, sweaty example of a solution looking for a problem. I'm skeptical of many claims people are making wrt how it's helping them.

My experience is it's like having access to an idiot-savant intern. Awful at most tasks, but knows everything and can read incredibly quickly.

Publicly, I've taken on the mantle of a staunch critic of generative AI and a pro-human, pro-soul advocate.

And for the most part, I'm happy with that stance. I like it. It feels good to rail against something, and it feels good to contrast a thing that I hate against something I love. It throws the love into more relief.

Yet, I don't want to lose any babies in that bathwater, and I don't want to lose my intellectual honesty in the neurochemical rush of fighting for a cause. So I'd love to explore the best use cases of LLMs that you all are actually using, and actually finding beneficial, life improving, productivity increasing, all of that.

I'd love to hear your experience, and ideally, you'd to tell me how you're doing what you're doing with it in enough detail so that I can try it.

I'll start.

Absolutely most useful thing I've found for it so far, and it's not even close, is language learning.

I'm in a slow process of learning Japanese, and asking a chatbot to break down the grammar of a specific sentence is super useful. It's also great for generating content for flashcards. Say you have a set of characters, and you want some example words that use each particular character. It's so easy to generate stuff like that.

Outside of that, I use it in super basic ways (basically as google with one less step).

So please, give me your best use cases, things that you've not only been impressed by, in a "oh wow, that monkey can tap dance!" way, but that has actually improved the quality of your life.

Please help me stay intellectually honest! I'm not a fan of generative AI in general, and LLM technology specifically. I think its capabilities are being drastically over-hyped. It's a perfect, sweaty example of a solution looking for a problem. I'm skeptical of many claims people are making wrt how it's helping them.

My experience is it's like having access to an idiot-savant intern. Awful at most tasks, but knows everything and can read incredibly quickly.

Publicly, I've taken on the mantle of a staunch critic of generative AI and a pro-human, pro-soul advocate.

And for the most part, I'm happy with that stance. I like it. It feels good to rail against something, and it feels good to contrast a thing that I hate against something I love. It throws the love into more relief.

Yet, I don't want to lose any babies in that bathwater, and I don't want to lose my intellectual honesty in the neurochemical rush of fighting for a cause. So I'd love to explore the best use cases of LLMs that you all are actually using, and actually finding beneficial, life improving, productivity increasing, all of that.

I'd love to hear your experience, and ideally, you'd to tell me how you're doing what you're doing with it in enough detail so that I can try it.

I'll start.

Absolutely most useful thing I've found for it so far, and it's not even close, is language learning.

I'm in a slow process of learning Japanese, and asking a chatbot to break down the grammar of a specific sentence is super useful. It's also great for generating content for flashcards. Say you have a set of characters, and you want some example words that use each particular character. It's so easy to generate stuff like that.

Outside of that, I use it in super basic ways (basically as google with one less step).

So please, give me your best use cases, things that you've not only been impressed by, in a "oh wow, that monkey can tap dance!" way, but that has actually improved the quality of your life.

Quick reflections on vibe-coding:

• kicks up weird addiction cycles in me
• the end product feels more like 20-30% jordan's creation and 70-80% LLM's. The text for example is all a little off, the style, etc, and it drifts because it's so easy to have it do it and so much more "time consuming" to go fix all the little bits (plus you have no guarantee it wont fuck something else up) and keep them coherent, versus just surrendering to the llm
• As a result I feel weirdly not that autonomous... like is the LLM working for me or am I working for it? This feels similar to the addiction cycle too
• There's a vibe that's similar to the "HR Lady" tone and em-dash hell of the LLM writing; like everything is sanitized and optimized for general professional sleeze marketing and you can't quite put your finger on why it sucks. The glassmorphism designs and gradients feel similar
• I continue to feel like I'm learning a lot from it, but I have updated to think that at the current stage its at, its actually an unhealthy practice and so is a little like a blood sacrifice that should be used sparingly. I imagine this will change, but maybe it'll get worse.

Checking where this post goes. Just checking if this ends up in a group

From the perspective of projection, what's Trump symbolize in you? Is it something in you that you haven’t loved yet?
Is it something in you that you haven’t boundaried appropriately?
What else?

Sometimes to do this kind of thought experiment i interpret my waking life as if it were a dream. When I look at how I view trump, what does he symbolize? And then I can ask how i relate to that to find out how I’m relating to that thing.

I will come back and do this self analysis in a moment.