lowsong

11 ms·

121.
by lowsong 1y ago
Unfortunately the Economist has shifted from "center right" to "far right" over the past few years, along with a general decline in reporting quality.
122.
by lowsong 1y ago
You are misrepresenting the situation. Users gave neutral questions and the generated response literally began praising Hitler.
123.
by lowsong 1y ago
I'd hope it was obvious I wasn't referring to the user. You shouldn't prosecute a person using a tool that wasn't intended for that function on themselves, in the same way you can't prosecute a person for performing
124.
by lowsong 1y ago
AI definitely isn't a bubble though, don't worry about it.
125.
by lowsong 1y ago
When the dust settles after the AI crash and people ask "how could we have known this was all so dangerous", I hope we remember this paper and remember the risks the authors took to try to stop this before it became the horror we
126.
by lowsong 1y ago
> The language of "generator that stochastically produces the next word" is just not very useful when you're talking about, e.g., an LLM that is answering complex world modeling questions or generating a creative story. Bu
127.
by lowsong 1y ago
Then it needs to be banned and prosecuted as medical malpractice. The "should" argument is very much not moot.
128.
by lowsong 1y ago
You don't have to see a game as "high art" to want to engage with its creators, their process, and the context in which the game was made. Interviews with game designers, game studio development blogs, developer commentary in
129.
by lowsong 1y ago
...but why? What's the point in playing a game that isn't an artistic expression, to communicate with another human? If there's no vision behind it, no expression to enjoy, then does it mean anything at all?
130.
by lowsong 1y ago
Therapy is one of the most dangerous applications you could imagine for an LLM. Exposing people who already have mental health issues, who are extremely vulnerable to manipulation or delusions to a machine that's designed to to produce
131.
by lowsong 1y ago
No the thousands of people working at Facebook did, he just got rich from it.
132.
by lowsong 1y ago
> Getting "dunked" only means it's controversial, not necessarily wrong. It undermines the author's position of being "moderate" if they align with perhaps the most decisive and aggressively written pro-AI p
133.
by lowsong 1y ago
> at the moment I’m mostly in tune with Thomas Ptacek’s My AI Skeptic Friends Are All Nuts. It’s long and (fortunately) well-written and I (mostly) find it hard to disagree with. Ptacek has spent the past week getting dunked on in public
134.
by lowsong 1y ago
I'm the first to admit that I'm an AI skeptic, but this goes way beyond my views about AI and is a fundamentally unsound idea. Let's assume that a hypothetical future AI is perfect. It will produce correct output 100% of the
135.
by lowsong 1y ago
Reading it back, my comment came off as more aggressive than I intended. I apologize for the tone.
136.
by lowsong 1y ago
> Instead of wrestling with Markdown formatting, Claude creates the document, helps formulate thoughts, and tests that everything displays correctly. Are... are you serious? Markdown formatting? The formatting standard that's design
137.
by lowsong 1y ago
Does no-one think ahead, for more than a moment? > Professional software developers are in the business of solving practical problems for people with code. Are we? Is that our job? You might think it is, and it might even feel like it. B
138.
by lowsong 1y ago
Compliance and security rules are the very last place you would want to trust an LLM to produce correct output. It's worse than not having security at all, as it can trick users into believing there are safeguards when they don't
139.
by lowsong 2y ago
I'm astounded that anyone even considered permitting LLM generated code in a browser engine, yet alone a technical committee. Why would you risk using a system that's known to generate logic errors and bugs in a system where sec
140.
by lowsong 2y ago
> … I believe our community should embrace it sooner rather than later — but like all tools and practices, with the right perspective and a measured approach. There is no such thing as a measured approach. You can either use LLM agents t
141.
by lowsong 2y ago
This is grossly misogynistic and insulting. It's the same old idea that women "just want to settle down and find a nice man" dressed up with some graphs. The author has even managed to make it about men in the end.
142.
by lowsong 2y ago
Nearly two years on from this article's publication, 'AI' tools are still as useless as ever for productive work.
143.
by lowsong 2y ago
> …it is pretty sad since this is not a matter of taste or political inclination: there must be a single truth This is more of a salient point that you perhaps realized. In life there is no single absolute, unknowable truth. Philosophy h
144.
by lowsong 2y ago
It's nice to see someone that gets it , and can explain why all of these generative-AI tools are completely pointless so well.
145.
by lowsong 2y ago
Sorry, I should have clarified. You can't do certificates that fast on Let's Encrypt no. I meant running a custom CA inside/alongside Kubernetes, and using that to issue 20-minute validity certs to pods.
146.
by lowsong 2y ago
I'm used to certs in Kubernetes, so even 6 days is long-lived. 20 minutes is more like it.
147.
by lowsong 2y ago
It's often very difficult to get domain names in large orgs, but very easy to get public IPs. An IP can be as easy as a couple of buttons to get a static IP and assign it to a cloud LB in AWS or Google Cloud. Domain Names usually requi
148.
by lowsong 2y ago
I've been in a weekly D&D game for over a decade. I don't keep playing it because of the game of D&D (certainly not, I could rant for a long time about the game's myriad shortcomings across many editions). I play it
149.
by lowsong 2y ago
I am baffled as to why anyone would want this, or anything like it. Sure, it's a cool technical demo. But... the point of D&D is a social game played with others. Even if you play a D&D-like video game such as Baldur's Gat
150.
by lowsong 2y ago
Last week I had to caution a junior engineer on my team to only use an LLM for the first pass, and never rely on the output unmoderated. They're fine as glorified autocomplete, fuzzy search, or other applications where accuracy isn&#x2

More ›