Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
cosmojg
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
14 ms
·
121.
▲
by
cosmojg
3y ago
1. Write prose in plaintext using preferred text editor 2. Add formatting using preferred markup language 3. ??? 4. Profit! Seriously, though, writing prose in a simple text editor and worrying about formatting later is far less distracting
122.
▲
by
cosmojg
3y ago
And Capybara be lookin' fiiine for tuning too. Seriously, though, you're right. These are some of the highest quality generative datasets in existence, and I'm surprised more isn't being done with them.
123.
▲
by
cosmojg
3y ago
It's about time the State started taxing the Church's land. Seriously, imagine if that money were invested in creating value instead of hoarding inelastic goods. We would all get to see just that much more prosperity.
124.
▲
by
cosmojg
3y ago
Whose pocket?
125.
▲
by
cosmojg
3y ago
I'm forever puzzled by the lack of rigorous coursework in probability and statistics in the American public school system, at all levels. I mean, heck, it's pretty much the only mathematical subject for which the question, "W
126.
▲
by
cosmojg
3y ago
Hot take, but if the FDA hadn't raised the bar for clinical trials to where it sits now (specifically with regard to the expenses associated with bureaucratic approval and regulatory compliance), I'd expect we'd see far fewer
127.
▲
by
cosmojg
3y ago
Listen, buddy, if I followed this advice, I wouldn't be able to use any software. As a matter of fact, Kdenlive is actually one of the better behaved tools in my toolbox. Nothing's perfect, but at least there are almost always r
128.
▲
by
cosmojg
3y ago
Lockfiles and version pinning are bandaid solutions to be applied in the absence of resources for supporting new versions of libraries, especially given that, unless you maintain your own internal package repositories, you're then depe
129.
▲
by
cosmojg
3y ago
The people who would be responsible for hosting such a Mastodon instance would likely be at least a bit more technically competent than those paid to post on Twitter.
130.
▲
by
cosmojg
3y ago
What's your preferred economic system, and what are its advantages and disadvantages compared to (neo)liberalism?
131.
▲
by
cosmojg
3y ago
I think it's a matter of setting expectations appropriately. LLMs aren't chatbots by default, they're text prediction engines. Depending on the use case, the instruct-tuned models can be quite a bit more difficult to use.
132.
▲
by
cosmojg
3y ago
From Wikipedia[1]: > Castilian (castellano), that is, Spanish, is the native language of the Castilians. Its origin is traditionally ascribed to an area south of the Cordillera Cantábrica, including the upper Ebro valley, in northern Spa
133.
▲
by
cosmojg
3y ago
> Alas, it takes longer to read blurbs :/ Tangential thought, but this sounds like it might make for an actually good application of large language models: reading blurbs (or entire books) and making recommendations accordingly.
134.
▲
by
cosmojg
3y ago
Not for GPT-4, unfortunately. Although, I'm certainly happy that Davinci et al remain available. I just wish they'd committed harder to what they had with code-davinci-002.
135.
▲
by
cosmojg
3y ago
I recommend reading the articles I linked as what you're saying is not true for most use cases. RLHF as implemented by OpenAI improves performance for one particular use case: chatbots. For every other use case, it degrades performance
136.
▲
by
cosmojg
3y ago
Which model would you recommend for a desk jockey who either runs or lifts weights a few days a week and goes hiking and climbing on weekends?
137.
▲
by
cosmojg
3y ago
I've had the opportunity to deploy LLMs for a variety of commercial use cases, and at least in these instances, I'd have to do something truly stupid for prompt injection to pose an actual threat to users (e.g., failing to isolate
138.
▲
by
cosmojg
3y ago
What was the point of moving away from the base model? I can't stop asking this question. Conversational formatting is achievable with careful prompting and a bit of good old-fashioned heuristic post-processing, and it was easier to ac
139.
▲
by
cosmojg
3y ago
As of today, Bard is now powered by the Gemini Pro model mentioned in the article. Bard Advanced is set for release early next year and will be powered by Gemini Ultra.
140.
▲
by
cosmojg
3y ago
Have you actually used federated social media? In my experience, it's more heavily moderated than any of the big centralized platforms. Like, to the point where one of the major reasons I still use the big centralized platforms is to g
141.
▲
by
cosmojg
3y ago
> By creating numbers from switches you can create even more interesting hot games. Well, don't leave us hanging! What are some of your favorite hot games on top of switches?
142.
▲
by
cosmojg
3y ago
> why did they go from internal politics -> external politics (large scale external politics) My guess is it has something to do with the hundreds of employees whose net worth is mostly tied up in OpenAI equity. It's hard to leve
143.
▲
by
cosmojg
3y ago
> Time preference theories of interest are unsatisfactory, because for them to make sense, lenders must have perfect knowledge about the future and be perfectly aware of competing usages of money at any point in the lending window. If th
144.
▲
by
cosmojg
3y ago
If someone's starving, I'll just gift them the money/food they need. I'm not going to set the expectation that they must return it to me. If I need to lie about the nature of the gift to ensure they'll accept it, so
145.
▲
by
cosmojg
3y ago
runpod.io is another good-and-cheap option
146.
▲
by
cosmojg
3y ago
Or buildings to humans!
147.
▲
by
cosmojg
3y ago
The paper[1] itself reads like marketing material that was itself written by GPT-4. Why was this published? And why is Nature reporting on it? As someone who generates fake/simulated datasets for a living, the scientific value of this
148.
▲
by
cosmojg
3y ago
The issue is that we consider passing peer review to be enough for something to be taken as truth when it should really be reproducibility/replicability. That, and the incentives currently driving academia are absolutely ridiculous. Pa
149.
▲
by
cosmojg
3y ago
Didya find it?
150.
▲
by
cosmojg
3y ago
Wow, yeah, that is strikingly similar! The video in the submission seems a little more dynamic, though, and a bit more relaxing.
More ›