

Then you can easily link the studies, right?


Then you can easily link the studies, right?


A lot of houses are owned by people who might no longer be employable (in that pictured scenario). Which then means they can’t afford to have their house fixed. Which in turn means those who live off of fixing houses get less contracts or have to massively reduce prices (and therefore dock their own income). So the potential changes are still a big issue, even if robotics aren’t as far yet.


No problem. If the society keeps up. If not, I wonder who will be able to buy anything which means the monetary value of AI goes to zero. Even most companies who could replace workers with AI only exist because they offer services that an AI can fully do (in that fictional scenario). So less companies, less money circulating, etc. What could go wrong.


Intelligence isn’t a one-dimensional axis either. Intelligence exists in many facets. Agents (so LLMs combined with tools, memory, etc.) overlap pretty much with one aspect that we perceive as intelligent. The main risk is basically the same as with humans of similar one-sided intellect: if you expect them to be equally intelligent in other areas, you will get disappointed or even hurt.


How senior are you? How much senior backing do you have? Close the PR with the comment “unreviewable”. Either they provider smaller chunks, or they can fuck off. If they are unable to prompt properly, that’s on them. They can use Claude all they want, but they should learn how to.


I think this is a real dilemma: do you contribute to an existing project, going through the motions of creating pull requests and having discussions or do you just make it yourself.
This might be surprisingly where a coding agent could motivate me. I typically don’t contribute, because in most of the cases I had the urge to get a missing feature into some piece of software, I had to sink countless hours into understanding their code, getting used to their style, bitching about the choices they made, and/or even setting up a dev environment to test the stuff in (which is often extremely badly (i.e. not at all) documented). This then quickly leads to not-invented-here-syndrome. With a coding agent however, almost all of this becomes irrelevant. The agent can figure out how to run it locally. It can figure out within seconds where in the code I need to make adjustments. If I don’t find an explanation for something in the code, I don’t have to search for ages, I can let the agent do the semantic search in a fraction of the time. Now I can focus on what I actually wanted to do: develop the feature, not doing code archeology for hours.


The US is so fucked in that regard. I think every other jurisdiction considers patent infringement only in regards to commercial use. A private person who makes no money from something shouldn’t be considered infringing on patents.


Isn’t the model part of the input to the algorithm?


Because the chatbot felt human enough to trick many people into a hype.




Citizens United
Such naming pisses me off so much; and that this kind of deception works so well.
In Germany we have for example the “Association of Family Businesses” who regularly advocate for libertarian shit. The name alone makes people think of the nice little baker two streets away that is run in the fourth generation, while the actual members are billionaires like Rossmann, Vorwerk, etc.
Oh or also the “farmer’s association” which typically lets people think of that nice old dude with his 10 cows and 2 tractors, while in reality it’s the big corporations (including dairy factories and chemical companies).


A lot of damage is still done, though. Just thinking of the fiasco in Finnland


That would actually be smart. With the current hype, there’s a lot of money floating around. Increasing that a bit more to fulfill regulations would be a non issue to them. We might even end up with more advances to green technology simply because at that scale it’s worth to invent new solutions.


Rather: they are free to do protocol negotiation with the DNS server I supply. If it speaks DoH, go for it.


You can spoof a lot with enough motivation. But it’s likely inconvenient enough that it will “catch” 90% of the targets.


Deezer is also pretty lenient with its family plan so far. However so is YouTube. YouTube family, which includes YouTube music, isn’t even excessively expensive if you actually share it between 6 people.


They are not that dumb. Apparently the code changes every few minutes (or seconds). Just a matter of time until they also check geolocation and/or IP during the scan.


I see. Yeah then it’s crap.
LLM workloads become faster with higher bandwidth, which directly translates to token per second. So it does matter for daily use.
But you are right that more and slightly slower is still better than less but fast.