zmmmmm

8 ms·

31.
by zmmmmm 2mo ago
It is also the case that the copyright holders are often putting restrictions around use of electronic forms that are driving the desire to use physical copies. I doubt AI companies would use a single physical book if they could avoid it -
32.
by zmmmmm 2mo ago
it's sort of non-controversial at the extreme end I think - where there is no human contribution, why should there be protection? But it just begs the actual question of how much human contribution there needs to be: - I wrote the
33.
by zmmmmm 2mo ago
but what if I really am a moron? how do I get that level of explanation now!
34.
by zmmmmm 2mo ago
energy / time / money / environment / office space reclaimed
35.
by zmmmmm 2mo ago
It's funny, because talk about a 25% potential productivity improvement saving multiple hours per day per employee in literally any other context and management would be crazy for it. Like AI hype bubble crazy. But culturally they can
36.
by zmmmmm 2mo ago
Let's just hope the pitch wasn't selling everybody's data for training
37.
by zmmmmm 2mo ago
> But do we still trust the open source software that is created now everything is vibe coded? I guess my question is, why would you trust it when it is not vibe coded? My suggestion is, have an AI review the code for you.
38.
by zmmmmm 2mo ago
TSMC in the US is a bit of a special case .... wouldn't surprise me at all if they were virtue signalling to the Trump admin more than anything with what they are doing there.
39.
by zmmmmm 2mo ago
I wonder if supply chain constraints currently represent an inherent scepticism of manufacturers that demand will last. It feels to me like a side effect of the self-dealing / incestuous financing that is going on is that manufacturers
40.
by zmmmmm 2mo ago
I don't want any of your arguments to be supported in law. The day someone can sue me because they searched for something on my web site and what they expected didn't come up, my web site is effectively legally prevented from havi
41.
by zmmmmm 2mo ago
Looking at the example where he asked for an SVG of a circle and it spent ages and drew a spectacular animated SVG with shading and a rotating arrow. It's honestly a bit concerning, I'm seeing this across the board (Opus 5, lookin
42.
by zmmmmm 2mo ago
It's a nice idea conceptually to decouple knowledge from reasoning but like everything in life I think it's something of a fantasy. When you ask a model to do something, its response is grounded in all the world knowledge it has f
43.
by zmmmmm 2mo ago
The question is, when is the "pro writer" version coming that lets you control this behaviour but costs more? Like night follows day, this will happen. They will need to dodge around the EU requirements but it will probably just c
44.
by zmmmmm 2mo ago
What I want is a version of `/clear` that keeps the conversation but drops out things like bloated logs, error traces, etc that were only relevant in the immediate local context. I guess compacting somewhat does that but I want somethi
45.
by zmmmmm 2mo ago
so i actually read beyond the TLDR and the reasons are: - it gets attached early so fully cached, even if later cache is busted - it gets included in every request automatically, so if your following requests are going to keep triggering Fi
46.
by zmmmmm 2mo ago
I'm with the author and others in this comment thread, speculating that effectively the balance has tipped to where humans are no longer the target audience of post training - other agents are. Whether it's through the reasoning &
47.
by zmmmmm 2mo ago
I think you lose a lot by not having the vision capability shared with the text. It is the joint reasoning across them where the power lies (the same model that sees the code and made the changes to produce the visual presentation, sees the
48.
by zmmmmm 2mo ago
Missing multimodal again? It is so valuable in practise to be able to have the models see screenshots - I guess if they aren't in the benchmarks then nobody will focus on it. But it completely nixes these for some of my main use cases.
49.
by zmmmmm 2mo ago
It's a combination of (1) and something you don't list: I think the frontier labs all have multiple generations of undisclosed models in continuous training. There is no "end point" when it's magically "ready&
50.
by zmmmmm 2mo ago
Some of the coming AI features are likely to be gated on how much RAM you have. So it may matter more in the future.
51.
by zmmmmm 2mo ago
> any better solution is either much more difficult or impossible. I just don't buy that. I think it is perfectly feasible to mandate device makers support parental controls that include content maturity settings. In many ways that
52.
by zmmmmm 2mo ago
the weird thing about it all is that age is a bad proxy in the first place for a child's level of safety in interacting with online content. Some kids are super mature at 11 and others still need protection at 16. For some reason the w
53.
by zmmmmm 2mo ago
the big thing containers don't allow is for the agent to run and use docker itself without compromising the host I'm not sure where "vast majority" cuts in but I would say a huge number of developers use docker and it is
54.
by zmmmmm 2mo ago
Meta knows how to win back developer's hearts .... let's see if they have the goods
55.
by zmmmmm 2mo ago
It feels to me that Anthropic is pushing hard here to try and establish acceptance of a tolerance of risk such that their tools can run fully automated without supervision, but directly on host hardware. I can see why they want that, but I&
56.
by zmmmmm 2mo ago
it's running a full VM so the agent can eg: run docker commands safely etc
57.
by zmmmmm 2mo ago
It's funny / ironic how the whole agentic development boom has taken off in terminal environments, but now the orchestration / IDE layer seems to be all Electron / GUI apps. Any suggestions for terminal UI versions of so
58.
by zmmmmm 2mo ago
> I definitely feel like there can be a pipeline where one extreme belief may lead to others Once you accept a premise that majority consensus of well informed people can be drastically wrong, it's not hard to then transfer that hor
59.
by zmmmmm 2mo ago
Tobacco company says "we are launching a new product that will cause cancer and kill people. It's highly addictive so we expect widespread uptake. We think it is crucial that regulation be introduced for mandatory regular cancer t
60.
by zmmmmm 2mo ago
> If anything, I want these models to be less persistent at their focus of completing their goal I think it's honestly a slightly ugly form of benchmaxxing - they are desperate to eke out the next few percentage points on completing

More ›