infinite_spin

3 ms·

31.
by infinite_spin 1mo ago
- https://openai.com/index/introducing-the-codex-app/ no where does this document suggest that codex can "one shot everything to perfection with just a prompt". It describes using a prompt plus agent ski
32.
by infinite_spin 1mo ago
Please stop dropping backhanded insults to other people here. It's not productive and violates guidelines.
33.
by infinite_spin 1mo ago
That's like comparing a self driving car to a bicycle. One is clearly more intelligent than the other.
34.
by infinite_spin 1mo ago
As pointed out in a previous comment of mine, it meets the definition of "intelligent", my last response is regarding the claim that I've been "fooled". Are we going in circles now?
35.
by infinite_spin 1mo ago
And yet this "next-token predictor" is able to churn out well tested, valuable solutions, to complex problems. If you want to downplay that as nothing more than a fancy auto-complete, be my guest, I lose nothing from that.
36.
by infinite_spin 1mo ago
> the ability to learn, understand, and make judgments or have opinions that are based on reason Agentic systems do this all the time. For example, I can point an agent at my codebase, and it will learn, understand, and make judgements b
37.
by infinite_spin 1mo ago
You might want to consider the difference between "lying" and "hallucinations", wherein one is shown that the agent knew it was being inaccurate, yet chose an answer that achieved some goal set forth; and where hallucina
38.
by infinite_spin 1mo ago
It's also a style that appears to deny the very first definition most dictionaries give for "intelligence" > the ability to acquire and apply knowledge and skills.
39.
by infinite_spin 1mo ago
The charges are the punishment. You are imprisoned, risk losing your career, risk losing reputation, and still have to hire a lawyer. The costs are incredibly asymmetric. This leads to the majority of people complying, which only emboldens
40.
by infinite_spin 1mo ago
Not to sound like a contrarian, but no. They would have thrown it in a faraday bag, cloned the device, and then would use one of the many existing forensic tools available to crack the pin. https://cellebrite.com/en/blo
41.
by infinite_spin 1mo ago
I feel like both viewpoints are valid ways of determining what you personally view as "good", and I don't want to take away from that diversity of opinion
42.
by infinite_spin 1mo ago
Number of successful outcomes reported (e.g. with disabled students, cancer patients). Labor costs. Latency of services. Scientific discoveries found. Etc. I'm not trying to discount the qualitative approach, but I think it's not
43.
by infinite_spin 1mo ago
How do you propose assessing "overall benefit of humanity" without metrics?
44.
by infinite_spin 1mo ago
I'm advocating for a litmus test that can't be easily abused, and I'm advocating against unfair equivalences. I also am not calling "overall a benefit to humanity" a metric, it's more of a conclusion we could a
45.
by infinite_spin 1mo ago
JSTOR was free for the public as well, but had a per-day download cap, which IIRC was 3 papers.
46.
by infinite_spin 1mo ago
That seems like an unfair equivalence. Lots of things you enjoy, including this very forum, benefit already wealthy private companies. That doesn't seem like it's a very good litmus test for whether something is overall a benefit
47.
by infinite_spin 1mo ago
I've found it's worse for simple tasks too, and I have to give it stricter guidelines, and sometimes it doesn't follow the same patterns I've grown to expect. I've found using 5.6 (sol) is good for diagnosing issues
48.
by infinite_spin 1mo ago
I'm not having that experience. So far each major model update has been at least slightly better than the last, in ways I've found useful. Can't say it's perfect, or able to do exactly what I want without a decent amount
49.
by infinite_spin 1mo ago
I have mine churning like butter and I'm rarely hitting a billion tokens per day, what's your workflow look like?
50.
by infinite_spin 1mo ago
https://archive.is/N9hmS
51.
by infinite_spin 1mo ago
The Hippy in me: "let's meet our new friends." The CEO in me: "I need hundreds of think pads."
52.
by infinite_spin 1mo ago
One of the earliest ad blockers I used would just replace ads with artwork. I'm not the author, but I found something similar here: https://github.com/jfonsecarivera/artblock
53.
by infinite_spin 1mo ago
> something real slow that takes up all their time Not to throw water on your plan, but the bots I've written intentionally run very slow with respect to each target. When done in parallel, across a wide range of targets, it doesn&#
54.
by infinite_spin 2mo ago
Welp, glad I posted this, because that's an extremely good point. Will have to think about this more. It kind of reminds me of how the reddit self harm notification system was abused. Maybe it can be an opt in type of system, or even b
55.
by infinite_spin 2mo ago
Consider these naming options: > total = calculate(items) > result = calculate(items) > value = calculate(items) > amount = calculate(items) All of them are reasonable options. If we bias the model's output so that one of t
56.
by infinite_spin 2mo ago
I think they justify it as an ease of use vs security trade off
57.
by infinite_spin 2mo ago
Fair point, we don't want to throw the baby out with the bathwater, however I am fearful of a surveillance state. These trust issues seem to be stacking up, so I think statutory transparency is necessary.
58.
by infinite_spin 2mo ago
At this point I think it's necessary to notify anyone who has been tracked after a fixed number of days, unless a court order is used to extend that period. An officer using this for legitimate reasons should have few problems getting
59.
by infinite_spin 2mo ago
My guess is it will be similar to how Genius watermarked lyrics, using things like variants of punctuation https://www.pcmag.com/news/genius-we-caught-google-red-hande...
60.
by infinite_spin 2mo ago
https://en.wikipedia.org/wiki/Buffy_(given_name)

More ›