zmmmmm

15 ms·

301.
by zmmmmm 8mo ago
and then the next: > you're hitting on the core tradeoff between correctness and usefulness The question is, is it a completely unsupervised bot or is a human in the loop. I kind of hope a human is not in the loop with it being such
302.
by zmmmmm 8mo ago
AGI is here it's just stupider than you thought it would be. Nobody really said how intelligent it would be. If it's generally stupid and smart in a few areas that's enough.
303.
by zmmmmm 8mo ago
a tale as old as time ...
304.
by zmmmmm 8mo ago
Just like the original OpenAI story, this seems like a case of reputation hacking through asymmetry in risk tolerance. There is not much novel about OpenClaw. Anybody could have thought of this or done it. The reason people have not release
305.
by zmmmmm 8mo ago
The lack of specificity to me suggests that someone probably is getting fired. It is written as if legal vetted every word.
306.
by zmmmmm 8mo ago
yeah, they should really have a think about how their behavior is harming their future prospects here. Just because you have infinite money to spend on training doesn't mean you should saturate the internet with bots looking for conten
307.
by zmmmmm 8mo ago
Dear news publications - if you aren't willing to accept an independent record of what you published, I can't accept your news. It's a critical piece of the framework that keeps you honest. I don't care if you allow AI s
308.
by zmmmmm 8mo ago
Especially direct quotes seems egregious - they are the most verifiable elements of LLM output. It doesn't make the overall problem much better because if it generates inaccurate discussion / context of real quotes it is probably
309.
by zmmmmm 8mo ago
> - There’s a difference. Users don’t see code, only its output. Writing is “the output”. So how is the "output" checked then? Part of the assumption of the necessity of code review in the first place is that we can't act
310.
by zmmmmm 8mo ago
I honestly don't really know. Notice they can't even actually say "human", they say "provider" above. So they are pretty much admitting, the alternative isn't that you get a life long relationship with a p
311.
by zmmmmm 8mo ago
The real healthcare system is rolling a dice too. A different dice, but an expensive one with a lot of delay and inconvenience - it's not like there's a choice that's problem free.
312.
by zmmmmm 8mo ago
> "AI can't read facial expressions, tone of voice, or body language," she said. "And those things matter. That's where the relationship between a patient and provider is built — in the nuance." It's am
313.
by zmmmmm 8mo ago
This should be a legitimate basis for legal action against whoever empowered the bot that did it. There's no other end point for this than human responsibility. Many of us have been expressing that it is not responsible to deploy tools
314.
by zmmmmm 8mo ago
I can't even get gemini on my phone, configured as my assistant, to schedule a timer. It just googles the answer now or tells me "Gemini can't do that". 16 years ago it was doing that perfectly.
315.
by zmmmmm 9mo ago
A system that tracks the integrity of each agent and knows as soon as it is tainted seems the right approach. With forking of LLM state you can maintain multiple states with different levels of trust and you can choose which leg gets remove
316.
by zmmmmm 9mo ago
It does beg, the question , whether any of this applies to less experienced people. I have a hunch that the open-ended nature of what can be achieved with AI will actually lead right back to needing frameworks, just as much as we do now, i
317.
by zmmmmm 9mo ago
I'm finding it quite a lot more assertive. It's doing things without asking every now and then. It cleaned up a whole lot of commented out of code that was unrelated to the change it was asked to make. Yes it's not great to h
318.
by zmmmmm 9mo ago
Let's savour this time pre-enshittification of AI. One day we will look back so fondly at it and wonder at our naivety.
319.
by zmmmmm 9mo ago
this concept of bottlenecking on code review is definitely a problem. Either you (a) don't review the code, (b) invest more resources in review or (c) hope that AI assistance in the review process increases efficiency there enough to k
320.
by zmmmmm 9mo ago
the hit to microsoft the other day was pretty interesting I saw reports attributing it to a miss on earnings from Azure but they were off by 0.4% on 39% growth. That's 39% instead of 39.4%. And the company stock dropped 10%. This is
321.
by zmmmmm 9mo ago
The whole site is fundamentally a security trainwreck, so the fact its database is exposed is really just a technical detail. The problem with this is really the fact it gives anybody the impression there is ANY safe way to implement some
322.
by zmmmmm 9mo ago
> brains that looked nearly a year younger Seems like a pretty small effect - if I'm 58 and I have the brain of a 57 year old, and to achieve that I did an entire year of exercise (as was done in this study) ... you'd have to e
323.
by zmmmmm 9mo ago
Somehow Apple maintains this perception that they make money on the hardware and therefore are trustworthy because they don't have any interests hostile to their users. But when you look at what's really happening it's clear
324.
by zmmmmm 9mo ago
It's not practical IMHO to move out of US owned cloud companies at this point in time. It could become so over years. However what is mostly practical is to avoid building on their proprietary layers and stick to relatively portable te
325.
by zmmmmm 9mo ago
> In the API, there are no changes at this time Curios where this is going to go. One of the big arguments for local models is we can't trust providers to maintain ongoing access the models you validated and put into production. Eve
326.
by zmmmmm 9mo ago
They compare it to software development but there is such a crucial difference to software development: by and large, software is an order of magnitude easier to verify than it is to create. By comparison, reviewing a vibe generated manusc
327.
by zmmmmm 9mo ago
I wish it could be more granular and ban the actual toxic elements rather than the concept as a whole. In general, people communicating with each other and forming social relationships is good. I get that it is hard to create granular laws
328.
by zmmmmm 9mo ago
very interesting! yes everything you say aligns with my experience and instincts.
329.
by zmmmmm 9mo ago
Curious what would be the most minimal reasonable hardware one would need to deploy this locally?
330.
by zmmmmm 9mo ago
I can believe in certain domains things can work this way, but we have way too many complex stakeholders and dependencies to have random deploys to production constantly happening without notice by anybody who feels like it. At very least t

More ›