ithkuil

5 ms·

61.
by ithkuil 4mo ago
We wouldn't have identities either if we were all clones and our memories could be edited and shuffled at each conversation. For an agent to have an identity we would have to intentionally make it hard to context engineering and limit
62.
by ithkuil 4mo ago
I assumed they are used to people being exhausted by those calls, so I was frankly surprised when she actually called again to complain
63.
by ithkuil 4mo ago
I guess the signal-to-noise ratio matters. In the last 3 months I received 700 spam/scam calls to my phone, my wife received about 400. We can't turn off ringing for unknown callers and we're getting mad. A few days ago I ven
64.
by ithkuil 4mo ago
I wonder if opus 4.8 would also be able to fix the code too
65.
by ithkuil 4mo ago
But LLM can write code that can do math and count. Tool use, more broadly, has proven to be a very powerful way to let LLMs do what they're good at (handle the fuzzy and imprecise nuances of natural language, which includes the scoopin
66.
by ithkuil 4mo ago
I'm honestly unsure if I'm more annoyed by slop or by the anyslop police at this point
67.
by ithkuil 4mo ago
Fair enough, there _could_ be powerful models that are hidden from the general public, but I wouldn't call it "naive" to think the current capitalistic incentives are such that the only way to produce such models is to do exa
68.
by ithkuil 4mo ago
An interesting way to rate limit access while also getting some data to analyze. They will lift this restriction later when they have more capacity
69.
by ithkuil 4mo ago
Why stop at bytes? Let's split it in individual bits and then look up the bits in pi! But Pi's binary expansion is not very practical for this purpose, since it's 11.0010... OTOH. e is 10.1011... Let's stick to fractiona
70.
by ithkuil 4mo ago
You'll find this an interesting watch: Reinventing Entropy Compression is Intelligence Part 1 3blue1brown https://youtu.be/l6DKRf-fAAM?is=ne73FCJ7ErXhzZ-v
71.
by ithkuil 4mo ago
Italy is no Denmark but you still require to register before selling you scrap copper. I think it's a reasonable response for a real problem and refusing to do this due to some idealistic free market principle appears to me to be a sig
72.
by ithkuil 4mo ago
Can an MCP provide prompts for your model to download and use CLIs (and ensure they have the right versions of those tools) in such a way that the data flows through the client side tools? The more I read this thread the more I'm convi
73.
by ithkuil 4mo ago
Possibly. But otoh one cannot complain that agents don't produce high quality code while at the same time not allowing them to thoroughly go through all the steps required to produce high quality code
74.
by ithkuil 5mo ago
How do you determine what's a fair price?
75.
by ithkuil 5mo ago
The informed claim is not that the obscurity layer has no value. Quite the contrary, it has such a great value that it basically reduces the incentives to have great proper security and thus once the obscurity layer is breached the second l
76.
by ithkuil 5mo ago
Even taking into account the fact that they are billing at 75% discount it's still quite cheaper
77.
by ithkuil 5mo ago
Aren't OCI caches content addressed?
78.
by ithkuil 5mo ago
Sparc (not a VLIW ISA) also had rotating register windows. But ia64 had a twist on it: the register window size was dynamic and "allocated" by the callee with an alloc instruction The only other ISA I know of that did something si
79.
by ithkuil 6mo ago
Is there evidence that frontier models at anthropic, openai or google or whatnot are not using comparable optimizations to draw down their coats and that their markup is just higher because they can?
80.
by ithkuil 6mo ago
That entirely depends on the location of the proxy and the extra conditions you can express. E.g. you could bind it to a source IP and have the proxy check that, or use some overlay network (like tailscale does) My point was that you don&#x
81.
by ithkuil 6mo ago
The auth between your app and the proxy can be scoped more easily. For example if the proxy runs in localhost you can trust the localhost workload. Or you can use some other kind of workload identity proof (like cloud based metadata servers
82.
by ithkuil 6mo ago
in exchange for service that presumably a) costs something to amazon to operate (so not pure 100B profit) and b) anthropic would have to spend anyway to operate their business. so basically ... you could view this as a kind of discount, but
83.
by ithkuil 6mo ago
Name checks out
84.
by ithkuil 6mo ago
If such tax would introduce an asymmetry that will favour human employment, then there should be enough buying power to create some demand
85.
by ithkuil 6mo ago
I wonder what percentage of the job space truly depends on the current edge we have over machines. I think it's reasonable to worry that way before machines are more reliable than the average human (let alone more reliable than a highl
86.
by ithkuil 6mo ago
Consequential enough that it's reasonable to plan for it.
87.
by ithkuil 6mo ago
Just train machines on the huge corpus of human needs so they can need things like no human has needed things before. What can possibly go wrong
88.
by ithkuil 6mo ago
Perfect place for the fortress of solitude!
89.
by ithkuil 6mo ago
Yeah. Basically conquest is possible when the victim is weakened. There are many ways to become weakened. Infighting and disease are common causes of weakening.
90.
by ithkuil 6mo ago
And indeed the Nahua people had writing. It's not about where you come from. It's about whether writing is useful or even required for some aspects of civilization to develop.

More ›