Also known as snooggums on midwest.social and kbin.social.
It is unpredictable because there are so many permutations. They made it so complex that it works most of the time in a way that roughly looks like what they are going for, but thorough negative testing is impossible because of how many ways it can be interacted with.
Not futuristic enough or something.
It isn’t very difficult, it is fucking impossible. There are far too many permutations to be manually countered.
D E E P E R
Or C: taking orders is easy and AI is on the verge of doing it better than a human any day now, just give us another billion dollars bro.
I doubt mobile orders will continue to grow as people still wait at the drive through because ordering while driving isn’t safe or practical. A big part of the convenience is just being able to stop by without a lot of preplanning.
Lazy authors of crime themed novels are sweating so heavily right now.
Project 2025: We have the rapture at home!
(25 years from now as per the article)
Anything 20 years or more away is a pipe dream that isn’t likely to happen anywhere close to speculation.
Personally I think that claim is a decent one: user-initiated request should not be subject to robot limitations, and are not the source of DDOS attack to web sites.
They are one of the sources!
The AI scraping when a user enters a prompt is DDOSing sites in addition to the scraping for training data that is DDOSing sites. These shitty companies are repeatedly slamming the same sites over and over again in the least efficient way because they are not using the scraped data from training when they process a user prompt that does a web search.
Scraping once extensively and scraping a bit less but far more frequently have similar impacts.
IT’S LOUD IN DA CLUB
They do it this way in case the data changed, similar to how a person would be viewing the current site. The training was for the basic understanding, the real time scraping is to account for changes.
It is also horribly inefficient and works like a small scale DDOS attack.
The AI doesn’t just do a web search and display a page, in grabs the search results and scrapes multiple pages far faster than a person could.
It doesn’t matter whether a human initiated it when the load on the website is far, far higher and more intrusive in a shorter period of time with AI compared to a human doing a web search and reading the cobtent themselves.
A user cannot physically generate as much traffic as a bot.
You are so close to getting it!
There is no difference between emptying a glass of water and draining swimming pool either if you ignore the total volume of water.
On the flip side, most websites are so ad-ridden these days a reader mode or other summary tool is almost required for normal browsing.
Firefox with uBlock Origin works perfectly fine and pages load faster without the ads!
But a user initiated operation isn’t the same as a bot.
Oh fuck off with that AI company propaganda.
The AI companies already overwhelmed sites to get training data and are repeating their shitty scraping practices when users interact with their AI. It’s the same fucking thing.
Web crawlers for search engines don’t scrape pages every time a user searches like AI does. Both web crawlers and scrapers are bots, and how a human initiates their operation, scheduled or not, doesn’t matter as much as the fact that they do things very differently and only one of the two respects robots.txt.
Or find a more efficient way to manage data, since their current approach is basically DDOSing the internet for training data and also for responding to user interactions.
AI crawlers tend to overwhelm websites by doing the least efficient scraping of data possible, basically DDOSing a huge portion of the internet. Perplexity already scraped the net for training data and is now hammering it inefficiently for searches.
Cloudflare is just trying to keep the bots from overwhelming everything.
The randomness itself isn’t the direct cause of the topic in the post though, because otherwise it wouldn’t be possible to reproduce the steps to get around any guardrails the system has.
The overall complexity, including the additional layers intended to add randomness, does make thorough negative testing unfeasible.