zmmmmm

8 ms·

151.
by zmmmmm 5mo ago
I use it on MacOS mainly, but I also switch it to my phone (Android) if I have to type something longer than a few words and it works well there.
152.
by zmmmmm 5mo ago
If you want a full size foldable keyboard with function keys and a trackpad, I can recommend the ProtoArc XK01[0]. This is a good enough keyboard that I use it as my main keyboard at home, and it's surprisingly cool to be able to just
153.
by zmmmmm 5mo ago
I have stuck with 4.6. I fully believe 4.7 can be smarter for truly complex and long running agentic use. But I prefer the more direct, literal mechanistic style and 4.6 seems to be peak Opus for that.
154.
by zmmmmm 5mo ago
I've always been fascinated that some people don't seem to have any email "voice" - they just can't translate email text into human emotional impact. So they write super abrupt emails, things they would never say in
155.
by zmmmmm 5mo ago
I will testify I have used V4 Pro as a coding agent and it did a great job solving a complex problem. It worked with Pi over something like an hour, iterating and running tests. I paid API rates via OpenRouter and it cost me less than $1 I
156.
by zmmmmm 5mo ago
it's honestly ironic that Java is removing its built in sandboxing right at the time when it would finally have been considered a high value feature. Between the need for AI agent sandboxing and the security apocalypse that is upon us
157.
by zmmmmm 5mo ago
once you use any language that lets you fluently inline a multiline lambda / closure you can never use Python again without it constantly irritating you
158.
by zmmmmm 5mo ago
As a side observation, it is striking but also not surprising in retrospect that the big successes in AI are coming from domains where things are fundamentally verifiable. Both software and math are either fully verifiable or low-cost verif
159.
by zmmmmm 5mo ago
So he's working on the singularity
160.
by zmmmmm 5mo ago
I have free electricity from solar and an old Macbook Pro M1 Max that has depreciated to zero and has no other use. Now how do the economics work out?
161.
by zmmmmm 5mo ago
I think AI rescue consulting is going to be come a significant mode of high value consulting, similar to specialists who come in to try and deal with a security breach or do data recovery. Purely AI written systems will scale to a point of
162.
by zmmmmm 5mo ago
> Why make your programmers wait? That depends on where the methodology goes. But more and more it's hands off. If the trajectory continues it won't matter because nobody is sitting their waiting / watching the LLM code an
163.
by zmmmmm 5mo ago
I'm very curious where we will saturate the curve on "enough" intelligence for coding. At some point, you can let a less smart model hammer at a problem for longer and get to the same result, and as long as you are not involv
164.
by zmmmmm 5mo ago
I've given V4 Pro some curly things and I was impressed at how it figured them out. I agree high level design is not its forte. But it sat in a loop and dogmatically debugged a crazy dependency issue to come to the right answer over th
165.
by zmmmmm 5mo ago
interestingly I had the same experience, and weirdly it's in part because it is clearly less intelligent. It's more of a mechanistic tool just doing what I ask (but still very smart and very competent about it) and less trying t
166.
by zmmmmm 5mo ago
HN always disappoints me with these kind of threads, with all the generic disappointment and Google scepticism dominating the conversation. Don't get me wrong, I'm still disappointed. But mainly because it looks so superficial. I
167.
by zmmmmm 5mo ago
Yeah i think this could backfire. At the moment they have such a clear messsage with Bedrock about data governance. You now have to ask a question and probalby get approval where previously there was no question and hence no barriers.
168.
by zmmmmm 5mo ago
yes it sounds like a hack to get access to untracked spend in corporate accounts. In my org, I have to file a form for reimbursement if I bought a pencil for $0.25 but in AWS? spend varies by +/- $5k per month and nobody even questions
169.
by zmmmmm 5mo ago
it's not going to help if you share a cache across security boundaries. That is what happened here and seems to be driving a spate of github action related problems.
170.
by zmmmmm 5mo ago
> Critic strength .... Sensor strength that's a nice breakdown I think there's something key you get at in terms of the combo of dynamic environment + type safety maximising both. With a dynamic environment, the LLM can do a lo
171.
by zmmmmm 5mo ago
history suggests so .... people do keep trying to make agent native tools and workflows, but time and time again it turns out to be better just to expose raw inputs and tools to them and let them work with those. See skills beating MCP in m
172.
by zmmmmm 5mo ago
I find it weird that the people steering Java have been seemingly willing to sit out the use case high performance computation while it has so dominated the computing landscape. They are just patiently incrementally iterating on all these J
173.
by zmmmmm 5mo ago
yeah we run it on a tiny setup compared to that and it's fine. In fact, impressively snappy if anything.
174.
by zmmmmm 5mo ago
Rather striking statements that have me somewhat concerned: > Agents open merge requests in parallel, trigger pipelines around the clock, and push commits at a rate no human team ever did. Git itself wasn't designed for that load, a
175.
by zmmmmm 5mo ago
I am curious the effect AI will have on these cloud offerings. On the one hand, they bust through a bunch of the pain points of setting it up and configuring it. Especially if you are trying to do it using something like Terraform etc. So t
176.
by zmmmmm 5mo ago
Interesting wording, because he's not the owner. What he owns is enough voting rights that nobody can challenge his decisions. And also interesting in the sense that, this is what he claimed to actually do a few years ago. He had a &
177.
by zmmmmm 5mo ago
It starts with the generic pitch around culture change - "culture eats strategy for breakfast" style. Then a bit of shock and awe around how extensively AI is going to redesign business processes in the long run , leading into an
178.
by zmmmmm 5mo ago
Ultimately yes, that was the issue. In theory they built a viable product, even if it still was cartoonish etc. But it was enough to see that even if it was perfected - there simply wasn't a killer app for what to actually do in th
179.
by zmmmmm 5mo ago
i think the key is, about half of that 33% can tolerate certain elements of it (stationary experiences etc) and another slice suffer in a way that will be resolvable or at least somewhat mitigated by technology improvements. And then anothe
180.
by zmmmmm 5mo ago
I can't tell how serious you are. But I'm curious - thinking of your past self (depending how old you are), what would have said about the current AI revolution 10 years ago? Eg: the chances that fully agentic generalised automat

More ›