xigency

9 ms·

91.
by xigency 2y ago
This is a topic I've been contemplating heavily and it's very instructive to see the iterative progress you've made here. I think development tools like this are necessary to expand coding outside of just the desktop experien
92.
by xigency 2y ago
Yes, if you view the footage of the bird strike on first approach you will see the landing gear is extended. ETA: The primary footage is hard to find now that the topic is so saturated, but there is a specific clip from a close vantage wher
93.
by xigency 2y ago
I agree with your take. Outside of how asocial and hostile society has become, there are certain people who understand more eccentric or fringe ideas and they can be quite rare. I can see how supplementing some amount of artificial conversa
94.
by xigency 2y ago
There are actually some algorithms specifically designed to optimize usage of cache resources without knowing the specific features of the cache. "Cache Oblivious algorithms" https://en.wikipedia.org/wiki/Cach
95.
by xigency 2y ago
> I didn't care to live in the Chicago suburbs, but I didn't mind working out there while living in downtown Chicago. I hear this a lot, but I'm quite the opposite. The suburbs are nice and quiet and I can go to the city f
96.
by xigency 2y ago
DRY... Don't Repeat Yourself DAMP... Do Always More Pasting
97.
by xigency 2y ago
Flashback to the days of doing system upgrades over Floppy Disk.
98.
by xigency 3y ago
It’s not unreasonable per se but seems a weak hypothesis without more supporting evidence.
99.
by xigency 3y ago
Alan Moore, Miracleman What a fantastic writer he is.
100.
by xigency 3y ago
Not every art movement in history has been motivated by awe. There’s a toilet in a gallery somewhere. Important to remember that these tools are extensions of humans and their expressions through time, and generative AI is still subject to
101.
by xigency 3y ago
I take the MIT license and excise the credit clause for publishing my least important personal projects. I feel it gives me just enough cover (no warranty etc.) and avoids requiring people to stick my name into whichever unaffiliated projec
102.
by xigency 3y ago
Aha. First comment I’m seeing addressing the super interesting fact that their methodology finds nonsense words to bypass constraints. I want to know how Sneaky Prompt finds these token embeddings and /why/ these specific tokens a
103.
by xigency 3y ago
You’ll catch more flies with honey than vinegar.
104.
by xigency 3y ago
> Well this post turned into a rant... You aren’t wrong. Frankly, it’s embarrassing. I could throw in a bunch more complaints and the kitchen sink but the point is we should expect better things from these companies and they should expec
105.
by xigency 3y ago
I don’t necessarily disagree insofar as for safety it is somewhat irrelevant whether an artificial agent is operating by its own will or a programmed will. The most effective safety is the most primitive: don’t connect the system to any lev
106.
by xigency 3y ago
I would be slightly more optimistic. They know each other quite well as well as how to work together to get big things done. Sometimes shit happens or someone makes a mistake. A simple apology can go a long way when it’s meant sincerely.
107.
by xigency 3y ago
Sam does believe in safety. He also knows that there is a first-mover advantage when it comes to setting societal expectations and that you can’t build safe AI by not building AI.
108.
by xigency 3y ago
The paper itself claims that OpenAI's actions have undone their stated goals: https://news.ycombinator.com/item?id=38374972 It has an excess amount of weasel words, so you might need to employ ChatGPT to read between t
109.
by xigency 3y ago
Yeah, that is a fair reality check for us terminal junkies.
110.
by xigency 3y ago
I just don’t know why an ethicist would remain involved in a company they find is behaving unethically and proceed with business as usual. I suppose the answer is the news from Friday, though the course feels quite unwise for the multitude
111.
by xigency 3y ago
Her position was listed as a fun fact, not in a responsible disclosure of possible conflicts of interest (though it ran the other way). Being at the top of the org and being present during the specific incidents that gives one qualms burden
112.
by xigency 3y ago
From the PDF: While the system card itself has been well received among researchers interested in understanding GPT-4’s risk profile, it appears to have been less successful as a broader signal of OpenAI’s commitment to safety. The reason f
113.
by xigency 3y ago
She is at the top of the pyramid. Did they not fire the chief executive? I am saying she is morally culpable for OpenAI’s actions as a controlling party. To put these claims in a published paper in such a naive way with no disclosure is aca
114.
by xigency 3y ago
> Considering what's in the charter, it seems like she didn't do anything wrong? It’s incredibly disingenuous to slap your name on an ethics paper claiming a company is doing malfeasance such as triggering a race-to-the-bottom
115.
by xigency 3y ago
As someone who has worked in this space before, using Blockly to help teach Arduino coding to kids, this looks fantastic! Keep up the good work.
116.
by xigency 3y ago
The organizations that run the most servers, at the largest scale, run Linux. It makes better operational and financial sense. But sure. The Mom and Pops of America still use Windows. (Who ever got fired for buying Windows?) Yet the backbon
117.
by xigency 3y ago
So I’m guessing you have never heard of AWS then…
118.
by xigency 3y ago
I use Faraday.dev on an RTX 3090 and smaller models on a 16gb M2 Mac and I’m able to have deep, insightful conversations with personal AI at my direction. I find the outputs of LLMs to be quite organic when they are given unique identities,
119.
by xigency 3y ago
Hm, it’s hard for me to say because most of my prompts would get me banned from OpenAI but I’ve gotten great results for specific tasks using finetuned quantized 30B models on my desktop and laptop. All things considered, it’s a better valu
120.
by xigency 3y ago
They could use Llama instead. OpenAI’s moat is very shallow. They’re still coasting on Google’s research papers.

More ›