> "the government made me spend 37% of my income on saving when I wanted to use it to raise kids."
This is a particularly funny one tbh. A nation's kids _are_ the retirement plan. It doesn't matter how many numbers you put in spreadsheets dated for 20-40 years into the future, if in said future, there isn't actually anyone to accept those numbers in exchange for labor.
In my experience this is what Claude 4.5 (and 4.6) basically does, depending on why its grepping it in the first place. It'll sample the header, do a line count, etc. This is because the agent can't backtrack mid-'try to read full file'. If you put the 50,000 lines into the context, they are now in the context.
> If you put the 50,000 lines into the context, they are now in the context.
And you can't revert back to a previous context, and then add in new context summarizing to something like "the file is too large" with how to filter "there are too many unrelated lines matching '...', so use grep"?
Using output-limiting stuff first won't tell you if you've limited too much. You should search again after changing something; and if you do search again then you need to remember which page you're on and how many there are. That's a bit more complex in my opinion, and agents don't handle that kind of complexity very well afaik.
> Cheat Engine doesn’t modify the binary. Ghidra can.
To clarify for other people who may not be familiar, (though I'm far from an expert on it myself) you can inject/modify asm of a running binary with CE. I'm not sure if there's a way to bake the changes to the exe permanently.
> Yes, but this assumes a finite amount of software that people and businesses need and want.
A lot of software exists because humans are needy and kinda incompetent, but we needed to enable to process data at scale? Like, would you build SAP as it is today, for LLMs?
Storage/compute/etc were orders of magnitude more expensive at the time, so the fact that it was 3-4 million is uh, pretty impressive? You could host a Matrix server for your 1,000 closest friends for basically no money.
You're absolutely right compute, network, and storage have continued to decrease in cost and accessibility.
The scale issue is enabling billions of consumers. It takes time and effort and skill.
It turns out that there are relatively skilled people who are willing to give their time and resources freely relative to billions of consumers in the market.
You know IRC isn't just one giant server serving every single user, right? Same for Mastodon. There were/are many different servers. Again, you are arguing against reality. IRCs/Forums have existed for decades, with hundreds of thousands of active users, with no problem whatsover. Scaling to billions is easy, since with more people using it, more people would be interested in hosting a server.
We've come full circle to banning advertising. It seems like we have good reason to believe that people will create the infrastructure for the communities that they _want_ to exist and fund them. So just banning advertising will probably be fine. Worst case scenario, we gradually loosen the ban. The advertising hellscape will grow back immediately, nothing of value will be lost.
Use 0.01% of brain power? How is it that Fox News always has the buy/sell gold ads? Hyper-segmenting society into advertising bubbles is about the same as if you hyper-segmented your body into cell clumps. You need unintentional cross-pollination, otherwise there is no more society.
Not necessarily. That's the old "I made Twitter in a weekend" joke.
That's not because you can technically replicate a product that your company will be successful. What makes a company successful are sales forces, internal processes and luck. Both are extremely difficult to replicate because sales forces are based on a human network you have to build, internal processes are either organic or kept secret, and luck can only be provoked by staying alive long enough, which means you need money.
I think something around that scale (say maybe 20 employees, but definitely not hundreds) was possible even before LLM got popular, but the people involved needed to be talented and focused. I'm not sure if AI will really change that though.
This is a particularly funny one tbh. A nation's kids _are_ the retirement plan. It doesn't matter how many numbers you put in spreadsheets dated for 20-40 years into the future, if in said future, there isn't actually anyone to accept those numbers in exchange for labor.