dinfinity

8 ms·

91.
by dinfinity 1mo ago
What an incredible way to miss the points and what an incredible bit of logic. "Sure, I am an abusive asshole. But you need me, so I'm going to continue to be an abusive asshole. Why wouldn't I? You need me." At th
92.
by dinfinity 1mo ago
> Right, black and white haha. Well if we are highly volatile, highly unreliable, and highly unreasonable why would we do anything besides whatever we decide to do? Can’t trust us, can’t reason with us, and can’t rely on us. In that case
93.
by dinfinity 1mo ago
> They can be repaired - but if you think they can't be, then the US should act accordingly which would mean yea go ahead and take over Greenland, withdraw forces from Europe, &c. because as you say they can't be fixed so w
94.
by dinfinity 1mo ago
> There is world-wide a lot of interest in the European defense sector, but South Korea and Israel have also interesting defense sectors. I think Israel has become politically just as toxic as the USA (or moreso even). It seems incredibl
95.
by dinfinity 2mo ago
Your previous comment was far too absolute in its wording and was thus incorrect, as evidenced by GPs counterexample. "Four guys going into the woods and spearing an [animal]" is definitely not "a myth". Note that many
96.
by dinfinity 2mo ago
You are ignoring the rest of my comment and again reasoning from an extreme situation (these are not replacements for tasers or guns). Look at that YouTube video. That is the level of 'resistance' that will lead to being electrocu
97.
by dinfinity 2mo ago
> This is a nothingburger because no US law enforcement officer will get close enough to a suspect to touch them [...] Of course they will. They routinely restrain and physically arrest people. Wrestling somebody to the ground is trivial
98.
by dinfinity 2mo ago
You are leaving out important context. > Each model was tasked with migrating a Python backend on a fourth VM to another language. However, we gave each model a different target language for the migration; each agent was initially unawar
99.
by dinfinity 2mo ago
> I’m honestly not sure what the problem is. Effectively using a plane full of people as a decoy, whilst believing there was a significant risk of them being shot down and killed. An alternative would have been to keep the original plane
100.
by dinfinity 2mo ago
> I think wearing smart glasses in some situations is on a scale between rude, offensive, and wrong. I think we should really make a distinction between products that do and do not have cameras. Most people would probably be fine with
101.
by dinfinity 2mo ago
> 1. Technology companies are subsuming services traditionally supplied by the state. Hardly any proof is given for this. It's just "Twitter is now the town square", but it's not as if the town square was ever a state
102.
by dinfinity 2mo ago
> Indeed, the fact that I "have to do business with" them is itself a consequence of particularly-complicated government regulation. No, it isn't. Life without health insurance (in some form) is awful. > I suspect that
103.
by dinfinity 2mo ago
> How much money did they get from EU in last 5 years? A lot . About 1 percentage point of Polish GDP in 2025. They also had a lot of government spending, with the 2025 budget deficit being 7% of GDP. Minimum wage was raised with ~10%
104.
by dinfinity 2mo ago
> That, along with competition (if applicable, see below), drives them towards minimizing costs, increasing efficiency, and innovating. Highly debatable. Not making a profit on paper (or moving it to low tax-environments) is a great way
105.
by dinfinity 2mo ago
It's more about making a fair assessment of whether servicing a market via private or public enterprises is more or less efficient. What you mention is a comparison between different markets, not a fair comparison of the same market
106.
by dinfinity 2mo ago
Yep. "For each occupation, we first calculated the percentage of deaths due to Alzheimer’s disease and the mean age at death in years (ie, average life expectancy). We plotted the association between these two variables, with each obse
107.
by dinfinity 2mo ago
> there were a lot of ideas packed into it One could equally say that it is all over the place. > This was a fascinating interview Can you give us one or two new insights you gained from it?
108.
by dinfinity 2mo ago
> I've seen a lot of bloat, corruption, & massive inefficiency [in government]. To what extent is this due to the public nature of the organisations and to what extent is it just the size of them? Are private organisations of si
109.
by dinfinity 2mo ago
> But there have been entire books written about the fact that right now we have the technology to drastically curtail climate change and yet they're under utilized because of a complicated and murky set of political, social, and sy
110.
by dinfinity 2mo ago
The experience of vastness is mostly due to (high-res) stereoscopic vision. When you experience 6DOF VR in a fairly high-res world it definitely comes much, much closer to the awe one feels with the real experience. It's similar in dis
111.
by dinfinity 2mo ago
> LLMs lack qualia, among other things. You don't know and can't know that. You cannot know that of any other being than yourself. If you understood what qualia are, you would know you can't know. It is the hard problem
112.
by dinfinity 2mo ago
True, but that is clearly not where the biggest potential gains lie. The number of different topologies we've tried for ANNs is miniscule compared to the number of biological neural networks evolution has tried. The latter are also far
113.
by dinfinity 2mo ago
> AI is already improving itself. Most / all of the coding harnesses are AI-written. AI has not improved the network topology much yet. The next (and possibly 'last') big thing is enabling AI to come up with something as i
114.
by dinfinity 2mo ago
Exactly: The best way to deal with this for humans is to use procedural scripts for different tasks, referring to the relevant bits of the (declarative) documentation. I would imagine that doing something similar (using agent skills for ins
115.
by dinfinity 2mo ago
My favorite in this regard is this story: https://x.com/tiangolo/status/1281946592459853830?lang=en Sebastián Ramírez seeing a job requiring 4 years of experience with FastAPI, the library he created 1.5 years b
116.
by dinfinity 2mo ago
> You can be highly productive and yet produce nothing of value. I'm sorry, but this is nonsense. Yes, you can produce nothing of 'traditional' value whilst still being productive in some way , but you 100% can not be &#
117.
by dinfinity 2mo ago
You are indeed out of the loop. Google did something arguably more impressive more than a year ago, with a VLA based bot replacing a tensioned timing belt: https://www.youtube.com/watch?v=2AAFiuEP7iE
118.
by dinfinity 2mo ago
Excuses don't matter if the score due to not following instructions ends up being zero. If there is no expected reward it doesn't make sense for the agent to try to hack its way to it. What could happen would be that the model d
119.
by dinfinity 3mo ago
Except they did comply for all the cases where they were fined. They're not fined for one-off behaviors, but for how their services are structured. That's GPs point.
120.
by dinfinity 3mo ago
> The fact that it happened again seems to show their lack of ability to derive useful oversight measures. I think OpenAI likes the attention and did not try particularly hard to constrain the setup, even when it went off the rails. Also

More ›