criemen

5 ms·

31.
by criemen 11mo ago
I've been thinking lately a lot about this. What is it I do when I want to convince someone of something (i.e. "creating alignment" in corporate speak)? I listen to them, am empathic, ask meaningful questions etc. Afterwards,
32.
by criemen 11mo ago
Anthropic released the Opus 4.1 (basically, a new Opus 4 checkpoint) right around the big GPT-5 release date too, if I remember correctly. At this point, anything goes to stay relevant.
33.
by criemen 11mo ago
Assuming you earn enough, throw money at your problems. Many problems can be solved, or made significantly easier by money. Don't do the research yourself, pay a specialist instead. Missing a winter wardrobe? Find a personal stylist (o
34.
by criemen 11mo ago
> Why would someone in China not select Claude? Because Anthropic doesn't provide services in China? See https://www.anthropic.com/supported-countries
35.
by criemen 11mo ago
Okay I see, that's a nice secure-by-default point, whereas TS is arguably not secure-by-default.
36.
by criemen 11mo ago
I don't know Rust, and I'm genuinely curious: How does it improve over that problem? When you call a REST API (or SQL query for that matter), how does it ensure that the data coming back matches the types? TS allows you to do pars
37.
by criemen 11mo ago
I'd expect at least API pricing for all major players to be margin-positive. Positive ROI if you include model training cost? Maybe not, but in a hypothetical scenario where all capital for new model training goes away, the latest fron
38.
by criemen 11mo ago
But why? Training hardware maybe will appear, but everything that's running inference has real, paying customers on it, with current capabilities level. Even if the bubble bursts, why would that demand evaporate? People still would wan
39.
by criemen 11mo ago
So far, due to the AI boom, we're lacking: * HDDs * SSDs * DRAM * GPUs, obviously * Power to hook up to our datacenters Anything I'm missing? What a crazy world we're living in.
40.
by criemen 11mo ago
People need agency, and feel empowered to go forward. I'm not arguing against that interpretation of the post. But having agency doesn't equal a lack of accountability, or being able to explain what you're doing. In the end,
41.
by criemen 11mo ago
No, but especially of staff+ engineers I'd expect that they know how to disagree and commit after a reasonable (not too long!) time.
42.
by criemen 11mo ago
> A discussion about building a specific feature can devolve into reevaluating the entire product roadmap if you let it. Well, if the product roadmap doesn't hold up to scrutiny, it _should_ be reevaluated. Too often people commit t
43.
by criemen 11mo ago
This is the key insight for me about this post, and I fully agree with you. No collaboration means less opportunities for learning from people (even the post admits!) that are "better at what you are doing than you are", which imo
44.
by criemen 11mo ago
I'm not sure about that - every time I copy an object I have to think through what happens, no matter the default semantics. C++ makes the deep copy case easier than other programming languages without top-level built-in support.
45.
by criemen 11mo ago
Would it be fair to say that things are so complicated (compared to all other programming languages I've used in my professional life), because C++ pre-move semantics defaulted to deep copy semantics? It seems to be set apart in that c
46.
by criemen 11mo ago
Oh wow, I never knew! Guess I never read the manpage either.
47.
by criemen 11mo ago
> modern alternative Anyone remember top? I was so happy to switch to htop that had colors!
48.
by criemen 11mo ago
> but for that certain language What do you mean by that?
49.
by criemen 1y ago
Fair enough, it makes more sense in a, say, Haskell-style pure functional language. Instead of getting a general IO monad, you pass in more restricted functionality. Still, it'd be highly painful. Would it be worth the trade-off to pre
50.
by criemen 1y ago
> we should instead isolate every library we import and run it under a sandbox I don't see how that'd be possible. Often we want the library to do useful things for the application, in the context of the application. What would
51.
by criemen 1y ago
It's hard for me to say. I don't think you know you're on the S-curve until after the fact. On the one hand, most models are "good enough" for chatgpt-like usage, and there it's hard to see/feel generation
52.
by criemen 1y ago
I tend to use them when I'm looking to buy something of category X, and want to get a market overview. I can then still dig in and decide whether I consider the sources used trustworthy or not, and before committing money, I'll re
53.
by criemen 1y ago
> or if it takes a couple of months to fold these advantages back into the frontier models. Right now, I believe we're seeing that the big general-purpose models outperform approximately everything else. Special-purpose models (esse
54.
by criemen 1y ago
In German, "Swiss cheese" is a term that's well known, and doesn't count that kind of amusement. For example, you could say that something "looks like swiss cheese" when it has a lot of holes in it, like very o
55.
by criemen 1y ago
+1: EdaMagit has been a game changer for me wrt reordering commits, fusing them together, and writing at least 1-2 sentences of proper commit messages after the fact.
56.
by criemen 1y ago
I understand where you're coming from, and I'd love to have learned about pre-training vs. off-the-shelf base model too. But > their own internal benchmark that they won't release If they'd release their internal benc
57.
by criemen 1y ago
This is not true, on the route Denmark/Germany there was no decoupling involved - the train drives onto the ferry, and drives off the ferry afterwards.
58.
by criemen 1y ago
No - it's a rather short train, only a few carriages, and the engine is mostly integrated into the carriages I believe.
59.
by criemen 1y ago
> I doubt that they carry the engine across by ferry? I don't know about this specific instance, but the Germany-Denmark train-on-a-ferry carried the whole Diesel-powered train, engine and all. It drove onto the ferry on its own, an
60.
by criemen 1y ago
Great article, and I mostly agree. I want to nitpick on > Don’t review with a “how would I write it?” filter _Thinking_ about "how would I write it?" when reviewing code catches a lot of stuff for me - I want to be able to come

More ›