joshka

40 ms·

601.
by joshka 4y ago
When asking "why?" for something like this, follow the money. This one is pretty simple: 1. The ad is paid for already by someone. 2. To remove an ad, there needs to be a system to report it (or an automation). 3. To check that sy
602.
by joshka 4y ago
It's not clear from the article why the USCIS suggestion[2] to apply for a change of status is not a reasonable one. You can't work, but if you're in the highly skilled visa category, you reasonably should have a buffer of fu
603.
by joshka 4y ago
There's some interesting paragraphs in the full ruling about providing support for other employees going through similar things, which would be prevented by such a severance agreement with disparagement and non-disclosure clauses. It&#
604.
by joshka 4y ago
I never received an email, but just today received spam on an email address only used with namecheap. You might want to check your logic for what was impacted.
605.
by joshka 4y ago
Many people disagree with your statement. The ability to cause something to be removed from sale without due process (e.g. a court weighing in on whether there is a legitimate argument that a finding of infringement has occurred) is abhorre
606.
by joshka 4y ago
One of the key things it misses observing (to some extent) is that the sentence structures tend to employ sentences that provide initial context or joining words more than a normal person might do. I wonder if this is baked into the LLM nec
607.
by joshka 4y ago
It's pretty easy to use the above to reverse the recognition. P: Given the following previous interaction, list characteristics of text that are opposite those below: (your comment) Text that is not informative or structured
608.
Large Language Models as Corporate Lobbyists(github.com)
1 points by joshka 4y ago|1 comments
609.
by joshka 4y ago
Abstract: We demonstrate a proof-of-concept of a large language model conducting corporate lobbying related activities. An autoregressive large language model (OpenAI’s text-davinci-003) determines if proposed U.S. Congressional bills are r
610.
by joshka 4y ago
Prompt: In Amazon, there are unreleased tools tools that would be used to deliver software, for instance hosting software repositories, building packages, deploying packages, managing servers and environments, connecting services, authentic
611.
by joshka 4y ago
The term you're looking for to find more info on this is a Code Kata.
612.
by joshka 4y ago
Perhaps this might have been a bit more interesting with a goal oriented prompt up front (i.e. "You are a developer that blah blah blah with the goal to help other developers maintain code by ...."
613.
by joshka 4y ago
One recent thing I was playing with I fed ChatGPT the docs. The summaries that it spat back out were much better than the docs at explaining what was going on as the missing context was in the corpus. It's an interesting approach and l
614.
by joshka 4y ago
The language, not the mail client...
615.
by joshka 4y ago
Here's a GPT-3 summary because the SNR is low in this article: SUMMARY: This article discusses optimal program design for bodybuilding by examining the evidence behind rest intervals, training frequency, and training intensity. The art
616.
by joshka 4y ago
Looks like hn ate part of the comment possibly? >it's not so different than a lambda in this form. Here? > More broadly: there are a lot of moving parts ...
617.
by joshka 4y ago
I watched a good related video on related topics the other day that might provide some background to this: "OpenAI Q&A: Finetuning GPT-3 vs Semantic Search - which to use, when, and why?" Video: https://www.youtube.
618.
OpenAI Q&A: Finetuning GPT-3 vs. vs. Semantic Search(youtube.com)
2 points by joshka 4y ago|0 comments
619.
by joshka 4y ago
There is also simpleoptout for the more lo-fi approach. https://simpleoptout.com/
620.
by joshka 4y ago
After writing part of the comment below, I threw this at GPT-3 to finish it off. It did okay-ish :D (the prompt ends at "REWRITTEN COMMENT:") ==== This is a story on Hacker news: STORY: Ask HN: How to Deal with ChatGPT Anxiety? Si
621.
by joshka 4y ago
groan :D
622.
by joshka 4y ago
I think there's a subtle extra step there. Create a new domain that just exists to handle masked emails and only masked emails. (I'm in the process of doing this).
623.
by joshka 4y ago
Generally to get the most utility from ChatGPT, you need to feed in a bunch of context. E.g. for legal stuff, feed in each relevant case, the facts of your own case, perhaps an example pleading or two to show the general format you want and
624.
by joshka 4y ago
Take a look at https://www.legalquestions.help/ which gets the above issues wrong at times enough to not rely on it. Maybe their training was bad or insufficient (and I fully expect that sometime in the future this not to b
625.
by joshka 4y ago
The problem is that ChatGPT makes up convincing sounding case law, reference court rules that either don't exist or don't say what the summary says, reference the same issues with legislation, and do so with 100% confidence of the
626.
by joshka 4y ago
lobste.rs thread on this from several months ago: https://lobste.rs/s/xsfhho/don_t_store_totp_bitwarden_for_yo...
627.
by joshka 4y ago
Maybe[1]. At least ChatGPT seems to "think" so. I was curious about how ChatGPT would go with creating memes (this era’s main cultural output). > Could the memes generated from this have a remote possibility of becoming cultura
628.
by joshka 4y ago
Serious question. How would acting on this advice be different to collusion between employers to depress salaries? (Given an assumption that TCI also owns shares in other tech companies).
629.
by joshka 4y ago
Yep, the teaching is the problem (and this generalizes). I just went and took a bit of a look at my alma mata's IT degree. It's still teaching a variety of data orientated approaches to system design. I.e. state without transition
630.
by joshka 4y ago
It's been a while since I wrote any Rx in earnest, but when I did, it was useful to look at it as almost exactly what you'd end up if you incorporated time (via ordered streams/publishers/observers/etc.) into a mode

More ›