Example: Patrick Stewart on Sesame Street https://www.youtube.com/watch?v=hA7lv1SDzno
70 karma · joined March 8, 2015
Yes I know they can escape containers, but that's not what I'm tasking them with.
Context is intuitively important, but people rarely put themselves in the LLM's shoes.
What would be eye-opening would be to create an LLM test system that periodically sends a turn to a human instead of the model. Would you do better than the LLM? What tools would you call at that moment, given only that context and no other knowledge? The way many of these systems are constructed, I'd wager it would be difficult for a human.
The agent can't decide what is safe to delete from memory because it's a sort of bystander at that moment. Someone else made the list it received, and someone else will get the list it writes. The logic that went into why the notes exist is lost. LLMs are living the Christopher Nolan film Memento.
I looked at Screenpipe but it doesn't fill me with confidence. The ideal solution would be a device-agnostic self-hosted service for archiving and searching screenshots. Would make data migration super simple. And then the app snapping the screenshots could be almost trivial.
Digital files have infinite supply. Draw a classic supply and demand diagram using that information and look at where the price should be. The supply curve is a vertical line at infinity on the x-axis. The demand curve crosses it at $0. Everyone knows and feels this intrinsically.
The artist is the thing with extremely limited supply. Something like Patreon will end up being the correct model.
A) Making the camera protrusion ungainly so that most people think "I'm going to put a case on this to make it look and feel better."
B) Changing the shape of the protrusion periodically to force people to get a new case.
Therefore adding an invisible extra $40-50 profit on millions of iphones sold.
Don't have to look much farther for motives. However, may I suggest a slight amendment to your statement and thinking:
The association with keybase.io does much to give the appearance of credibility, trustworthiness and stability to cryptocurrency.
I've always thought it best to just cut straight to the core of issues like he's doing. I suppose it's a jerk move to corner people into admitting they're wrong even when it's obvious. Maybe there's a nicer jedi-mind way to do it.
You order a burger at the cashier window, then go to the pickup window. If the burger is already made, it's already at the pickup window when you get there.
The author wants a special case where if the burger is already made, they hand it to you immediately at the cashier window. This might seem more efficient, but both in the restaurant and in code it makes logic way more complex.
Getting "stopped out" is fine, it's part of investing. You need to stop your losses. I'm just saying don't put the order on the books for the exchange to pluck like a ripe berry.
Stop-losses are very important in investing but I'd recommend always executing them manually at market rate. Set a price alert to notify you. Of course if the exchanges are down you're sol, but if BTC is on a bullet train to $10 your exchange stop-loss probably won't fill at your price anyway.
ZFS/Raid + ECC ram are complicated and expensive and are not even a 100% guarantee against bit rot.
Sometimes I feel bad complaining about this. Things are evolving fast for a reason. Everyone is trying to make every part of this process better. The backbone joke above illustrates the point. Would you go back in time and freeze the evolution just to regain sanity for today's devs?
How does this handle dupes and near-dupes? People have photos scattered all over. Can I throw a random old folder of photos at this and trust that it won't overwrite an otherwise identical jpg with a bit-rot corrupt one for example?
You mention idempotence in your blog post, if I run this twice using the same 'unorganized' source folder, will it copy all the files again the second time?