kay_o

5 ms·

1.
by kay_o 7d ago
I am not saying anything is ok, where did you read that?
2.
by kay_o 7d ago
> There's no reason for that person to be allowed to operate a vehicle Doesn't much of america outside of cities also have the secondary problem of you physically cannot feed yourself, get to appointments, or do anything withou
3.
by kay_o 9d ago
Imho part of isuse here is that you cannot "partial AV" it. You cannot have a car that will EVER need you to intervene or else it does not work because of awful human nature. The amount of tesla uber drivers ive had literally brow
4.
by kay_o 10d ago
Threat model a bit different, they can just request xcancel.com/unique-string and check what IP or account requested unique-string IMO
5.
by kay_o 11d ago
What about this can't be resolved by a PAC-file?
6.
by kay_o 14d ago
Probely because that is the entirety of it. There is zero story or creativity or anything else that goes into an actual game, virtually nothing is named and any that has is a simply word or two jammed together into a name -- the player guid
7.
by kay_o 15d ago
If you are caught midway, you get banned for abuse/ToS, and you are not allowed to delete/edit your comments-account if you are unable to sign in. (I am guessing they are retaining data under some bullshit about gdpr legitimate in
8.
by kay_o 15d ago
When doing basic CRUD apps I can count on fingers the amount of times guard rails haven't tripped and ended the session
9.
by kay_o 16d ago
It's usually wrong.
10.
by kay_o 17d ago
In this specific case, they are already trespassing to be able to read the sign against the owner's wishes, are they not? The sign is only placed in a hidden corner visible by trespassers and not visible to any regular customer.
11.
by kay_o 17d ago
com net org org has always been owned and managed by the US
12.
by kay_o 17d ago
It's not malware. You chose to read their site.
13.
by kay_o 18d ago
Is this work related or addiction related? You are repeatedly pulling a slot machine in hopes for a jackpot of working usable code, something which does not come up just often enough to require another pull of the conversation turn. Claude
14.
by kay_o 19d ago
I can't tell if they're trying to fix it with slop now but clicking the number of reviews now jumps below the description. To nothing. Things are going great
15.
by kay_o 20d ago
Presume this is an experiment or phased rollout. I was just on Amazon and only the first page of reviews are visible if logged in and no reviews are visible if logged out. edit: It seems badly done too as the (140) count next to the stars
16.
by kay_o 20d ago
> I'm not asking it to rewrite. I'm asking it to fix my spelling, No matter what you demand, LLM will rewritten, no matter how you tell it not to, it will be rewritten. Your only option is to use traditional corrections such a
17.
by kay_o 23d ago
I don't even need to tiptoe ! Not being there and not prompting anything is enough to trigger safeguards. Having not asked a single security question it will write wildly vulnerable code, go back and fix it, and guardrail itself out of
18.
by kay_o 26d ago
> They seemed to suggest they were running it on their own hardware? Where did you see this? I was to manually opt in to "allow models hosted in China" and there is writing about how their no data retention agreement with DeepS
19.
by kay_o 26d ago
What plan do you use at Fireworks? I've been rugpulled multiple times on their Fire Pass subscription
20.
by kay_o 29d ago
It definitely isn't but Claude for some very unique reason enjoys to overthinking and go on side quests in the stupid ways I've not seen Codex, DS, Kimi, Mistral do at equivalent effort and thinking setting.
21.
by kay_o 1mo ago
I use mailboxorg its acceptable. The webmail has io.ox in the URL so I assume it is hosted ox (www.ox.io of softwares) which is SIGNIFICANTLY worse performance and usability wise compared to the Fastmail webmail -- I don`t use SMTP/IMA
22.
by kay_o 1mo ago
This is what every chinese "credit reseller" is doing, selling all the prompts and responses to labs in addition to selling quanted kimi/ds/etc as opus/fable
23.
by kay_o 1mo ago
AI slop and you do not contribute other than self promotion. Yes it is promotion.
24.
by kay_o 1mo ago
My largest problem with H1 is how braindead scripted/AI their triage is. - Starting scenario: no way to contact a company outside of H1 (or some other managed programme) - The company is compromised, their customer support has no idea
25.
by kay_o 2mo ago
It is here: https://radar.cloudflare.com/scan/08d947f4-488e-486a-abfc-42... It is also flagged on urlhaus.abuse.ch and a few others
26.
by kay_o 2mo ago
Multiple (major) services are detecting opencode's domain as malicious. Presumably SFR uses those feeds. Check virustotal it the URL
27.
by kay_o 2mo ago
Multiple services are detecting opencode's domain as malicious. Cloudflare uses those feeds. Check virustotal it the URL
28.
by kay_o 2mo ago
You are a bad user to them. You are not a high value user that is valuable, registering with your real, non throwaway email that consumes meme slop and stolen content without ad blockers.
29.
by kay_o 2mo ago
It is nothing. There are many individuals that run 10 seats, not even business, and unless you are paying API per token rates you are a loss already without answering a support ticket ! Unfortunately ...
30.
by kay_o 2mo ago
This was serious post sadly ! All self serve tiers Claude Team, OpenAI Business, ... all are considered the same as personal plan. The only differentiated tier is [self service] || [enterprise support contract with guarantees]

More ›