gabosarmiento

Thank you for sharing this. Great job.

Because if the result is just more AI slop people will stop caring.

This reminds me of the ads in Idiocracy. I really hope this isn’t the direction advertising is heading.

Have you consider asking your agent to set up basic permission deny rules to block dangerous commands? Just ask it and request that it shows you the config before applying it.

Lol that’s a great idea, basically the Million Dollar Homepage, but on a car

How do you prioritize what to focus on and what to reject? I mean what does your prioritization framework look like? what's your margin of decision? And 2nd or 3rd question if I may, how do you stay up to date? Like what sources do you use or do you mostly rely on AI?

I would like to see how each recipe performs against its corresponding evals. Some sort of ranking would be useful.

Everyone keeps posting articles about how to implement RAG, but I also wonder why there isn’t some sort of skill to help people create a simple retrieval plan, starting with the retrieval methods and connecting them with evals. This could show whether they actually improve the result and make retrieval simpler for any agent, instead of making people start from zero.

I don't know, I personally wouldn't buy it as it is today. I'd focus much more on memories now than the afterlife.

The critical part for me is trusting you with all my info. But once I cross that chasm, I'd talk to it on whatsapp, send photos, random memories, little stories, and years later I can actually interact with that part of my life instead of just scrolling photos.

And yes, you can still have the account live after you die, but that feels more like a feature than the core product.

I guess the moat is making each person's memory archive so valuable with the AI assistant that you simply don't want to move it.

Thanks for sharing. Made me remember Formspring, but now anonymous people submit things they actually told/asked AI, creating a very interesting(and probably quite addictive) public feed of how humans are really using AI.

AI;DR This is the broader symptom of a global toxic corporate culture.

We are afraid that one badly phrased message can hurt our reputation or even our jobs, so optimizing every message with softer wording, more political correctness, less directness is safer.

But the real problem is the culture underneath it. If people are constantly worried about how every sentence will be interpreted, eventually they stop speaking plainly, stop challenging decisions, and sometimes stop speaking up at all. So this is like attacking the consequences not the root causes. There should be some sort of measurement of how broken is a community or a company by letting a bot scan Slack/Teams and let the bot remind everyone the importance of lowering the fake index.

I’d think about this less as “how do I market this?” and more as “how do I find a repeatable acquisition machine?”

Do you know roughly what a customer is worth to you? If a customer is worth $1,000 and you’re willing to pay $100 CAC, then a channel you expect to bring 10 customers gets a $1,000 test budget.

Run the test. If it gets close to that CAC, put more into it and see if it scales. If it doesn’t, kill it and test another channel.

Once you find the one that works, that’s what you accelerate.