russian dude generates botslop on dating apps, makes 5000+ girls talk to bot, chooses the one that stayed with chatgpt: https://xcancel.com/biblikz/status/1752335415812501757#m
this post was submitted on 28 Oct 2024
1 points (100.0% liked)
TechTakes
1432 readers
16 users here now
Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
founded 1 year ago
MODERATORS
In separate investigations completed by the blockchain firms Chaos Labs and Inca Digital and shared exclusively with Fortune, analysts found that Polymarket activity exhibited signs of wash trading, a form of market manipulation where shares are bought and sold, often simultaneously and repeatedly, to create a false impression of volume and activity. Chaos Labs found that wash trading constituted around one-third of trading volume on Polymarket’s presidential market, while Inca Digital found that a “significant portion of the volume” on the market could be attributed to potential wash trading, according to its report.
load more comments
(2 replies)
Posting not for the content but for the stunningly inept AI slop illustrations:
https://datastream.substack.com/p/mistakes-from-my-failed-startup-in
Amazon used an AI-generated image as a cover for 1922's Nosferatu, and it got publicly torn apart on Twitter:
On a personal note, it feels to me like any use of AI, regardless of context, is gonna be treated as a public slight against artists, if not art as a concept going forward. Arguably, it already has been treated that way for a while.
You want me to point to a high-profile example of this kinda thing, I'd say Eagan Tilghman provided a textbook example a year ago, after his Scooby Doo/FNAF fan crossover (a VA redub came out a year later BTW) accidentally ignited a major controversy over AI and nearly got him blacklisted from animation.
I specifically bring this up because Tilghman wasn't some random CEO or big-name animator - he was just some random college student making a non-profit passion project with basically zero budget or connections. It speaks volumes about how artists view AI that even someone like him got raked over the coals for using it.
What do they mean by "in color"? If it's just various tints throughout the film that's normal and cool. If they mean full on colourised that's messed up.
ffs it's in public domain just use a still from the staircase silhouette like everyone else
load more comments
(1 replies)
today i found out that openai wanted to make their own chips https://neuters.de/technology/artificial-intelligence/openai-builds-first-chip-with-broadcom-tsmc-scales-back-foundry-ambition-2024-10-29/
can you imagine what kind of disaster would it be
that article misses one of the delicious parts of that story: they called saltman a “podcast bro” in derision
OpenAI considered building everything in-house and raising capital for an expensive plan to build a network of factories known as "foundries" for chip manufacturing.
Oh man, that's a delicious understatement. If the allegations are true, this was a plan that would make the military-industrial complex envious.
Had a first-hand AI encounter today at the grocery store. The self-checkout now has a script that monitors an overhead video feed to make sure you're not getting tricky about what scanned and what got put into the bagging area, and if it thinks you're shady it will stop you from proceeding and summon an employee with no notification that something is wrong.
The new self-checkout process is as follows:
- Scan your item
- Hold the item plainly before you so the overhead camera doesn't get confused, looking like a Catholic priest about to deliver communion.
- Place item in bagging area. Try not to have to shift things around to find a place.
- Swear as the nom-mutable voice instructions tell you to bag "your... Item." Legitimately feels like they got as far as assembling the voice lines before anyone realised that having the compu-checker read every purchase out loud would lead to at best an unworkable cacophony if not several immediate lawsuits.
- GOTO 1
Even as antisocial and impatient as I am I've found self-checkout to be a UX disaster, but somehow it keeps getting worse.
sometimes i manage to confuse self-checkout overhead camera by having a bike helmet on, when that happens i have to hold it up over bagging area (but not put inside because weight won't match)
load more comments
(1 replies)
I call this the law of conservation of complexity
a quick interest check: I kind of want to use our deployment’s spare capacity to host an invite-only WriteFreely instance where our regulars can host longer form articles
…but WriteFreely’s UI is so sub-optimal the official instance (write.as) runs a proprietary fork with a lot of the jank removed, and I don’t really consider WF to be production ready out of the box.
we can point the WF backend at arbitrary directories for its templates, page definitions, and static assets though, so maybe I could host those on codeberg and do a CI job that’d pull main every time it updates so we could collaboratively improve WF’s frontend? it’s not a job I want to take on alone (our main instance needs to take priority), but a community-run WF instance would be pretty unique
the pros of doing this are that WriteFreely at least seems to have very slim resource requirements and it’ll at least reliably host long form Markdown on the web
the downsides are again, it’s janky as fuck (it only supports Mailgun of all things for email, but if you disable that the frontend will still claim it can send password reset emails… but it’ll check the config and display an error if you click the reset link??? but they could have just hidden the reset UI entirely with the same logic???? also I don’t like the editing experience), and it’s not really what I’d consider federated — it shoots an Article into ActivityPub whenever you post, but it’s one-way so replies, boosts, and favorites won’t show up from ActivityPub which makes it feel a bit pointless. there might be a frontend-only way to link a blog post to the Mastodon or Lemmy thread it’s associated with on another instance though, which would allow for a type of comment system? but I haven’t looked much into it. write.as just has a separate proprietary service for comments that nobody else can use.
this definitely won’t replace Wordpress but does it sound like an interesting project to take on?
@self If you fork/make a custom version of the software will you call it WriteAwfully
WritePhilthily
Love it.
manuals.awful.systems
also, what’s a funny subdomain for this kind of thing?
gibberish.awful.systems, to form GAS
mouthful.awful.systems
scribes.of
damn
I want that album
Sundar Pichal, Google Q3 2024 earnings call:
We're also using AI internally to improve our coding processes, which is boosting productivity and efficiency. Today, more than a quarter of all new code at Google is generated by AI, then reviewed and accepted by engineers. This helps our engineers do more and move faster.
Firstly, if this is literally true they're completely fucking cooked.
Secondly, if it isn't, what version of it is?
Best case scenario they are using a loose definition of AI to mean any code generated by other code in order to signal to investors that google isn’t the hulking, sluggish monolith that it is and is agile enough to use AI.
Worst case scenario: “hey chatgpt pls write me new search algorithm to print money, thanks, sundar”
from someone on Mastodon:
Google has a gigantic code generation culture, because the engineers there strongly prefer complexity to drudgery.
If you asked them to write fizzbuzz and left them in a room for twelve hours they would deliver a new programming language that generalized repetitive string printing, with an extension language for potential non-string-printing actions.
I left in ‘22 but feel fairly confident that “25% of code generated by AI” is going to be more of the same.
that's quite appealing to me ngl
God knows I like a good DSL, but "complexity over drudgery" just sounds miserable. I also wonder what kind of stuff they're coding that's supposedly trivial enough to be generated by AI.
Boss needs you to find the contiguous subarray with the maximum sum. Says he needs it by EOB Friday.
Here ya go boss, a 80% prototype solution.
/* TODO: support other element types */
unsigned int * maxsumsubarr(unsigned int arr[], size_t len, size_t * sublen) {
*sublen = len;
return arr;
}
I half want to jest "PDD strikes again" but honestly it feels like only half the explanation
(promotion driven dev)
Man now I’m thinking about AI written PIPs. God if I got an AI PIP I’d self immolate on company grounds.
oh this is almost definitely real, given that the regular PIP process was already designed to get you to quit
Firstly, if this is literally true they’re completely fucking cooked.
I totally believe it. Y'all remember Stadia? That was a cosmic freebie and Google absolutely dropped the ball on it so laughably hard.
which part of it was the freebie? whole service looked dead on arrival to me (for the simple reason of physics)
At least for me in the US, performance was very good. I was able to 100% Sekiro, for example.
The reason I think it was a freebie is:
- Everyone was stuck in-doors about six months after launch
- Everybody wanted to play videogames, but no one could get GPUs and the console situation was not great
- Cyberpunk ~~2022~~ 2077 came out and tons of people wanted to play it. It ran terribly on consoles and on PCs, but surprisingly well on Stadia at launch
It may have still failed altogether anyway, but the fact that they didn't seize this opportunity, and instead stuck by their absolutely confusing-as-fuck "like Netflix but not really; first let me explain how this works" subscription model, always gets me.
Edit: Cyberpunk 2077 🤦🏻♂️
but the fact that they didn’t seize this opportunity
honestly, I think they did try, and ran into the unfortunate reality of physics
to make that product work, you need reliable high throughput (this is helped by codecs), sufficient juggle-able GPU space (this is helped by being a gear-hogging first-in-line monopolist), and lastly the casual little requirement of actually being close enough to your customer base
iirc US cost to coast latency is around 65~70ms (so 2x that is the upper timebound for player interactivity, obvs there it'd be less because more local DCs though). just from me to europe is 165msec, with a far less predictable path throughput. the scale economics to launch a DC for this in ZA (even to serve subsaharan africa all the way up to kenya) just plain doesn't work, and there are many more places in the world where it doesn't
it'll be interesting to see if a retrospective as to why it failed leaks out of that biz someday
load more comments
(2 replies)
for the love of beebo can i just get like eleventeen seconds pls where i dont have to put up with the sociopathy that is academic cs