If you don’t have a Twitter account but like to search the network for the latest news, you can forget about it now. Out of sheer paranoia or genuine danger of data scraping by AI startups, Elon Musk has made moves no one would ever dream about.
It would never occur to anyone that, to read more than 600 posts a day, they would have to pay $8/month for Twitter Blue “verified” access. And even then, reading tweets wouldn’t be unlimited – it would only increase to 6,000 posts per day.
What on earth is going on?
Over the weekend, anyone who tried to access Twitter without logging in was welcomed with a message “Something went wrong. Try reloading.” And no matter how many a time you tried to reload, you would keep getting the same message.
This was the consequence of Twitter’s (or better to say Elon Musk’s) decision to deny access to anyone who wasn’t logged in. The reason behind such a decision was the alleged attempt of “several hundred organizations (maybe more)” to scrap Twitter data, “to the point where it was affecting the real user experience.”
However, besides banning access to unregistered users, Musk went a step further:
This seemed to have been the straw that broke the camel’s back. While for Musk it was all fun and games, Twitter users were enraged. Some saw this decision as an attempt to make everyone pay for the verified account, while for others it was the freedom of speech restriction.
These figures remained unchanged for a few hours. Verified users were able to read 6,000 posts per day, unverified 600, and new unverified 300. However, the numbers were soon to be changed.
The consequences and the response that followed
These technical issues on the social media platform caused a significant number of users to encounter error messages, such as “Sorry, you are rate limited. Please wait a few moments then try again,” or notifications stating that the site was unable to retrieve tweets. The reports of these glitches began at approximately 8 am EST and worsened as the day progressed.
As a result of these problems, individuals took to Twitter to express their frustration, resulting in hashtags like “Rate Limit Exceeded” and “#TwitterDown” becoming the top trending topics in the United States. The first hashtag had already been shared over 40,000 times by noon on Saturday.
Elon Musk responded to the situation by announcing that he would soon increase the tweet limits for different types of accounts. He gave verified users limit of 8,000 tweets per day, unverified users 800 tweets, and new unverified accounts would have a limit of 400 tweets. It appeared that as more complaints accumulated, Musk continued to raise the quotas.
The imposition of these new limits sparked widespread frustration among users, leading to the emergence of trending topics like “Wtf Twitter” and “Thanks Elon.”
At 6 pm EST on Saturday, Musk tweeted that the new quotas were now set at 10,000 tweets per day for verified, 1,000 tweets for unverified, and 500 tweets for new unverified accounts.
What is data scraping?
The exact meaning of Elon Musk’s reference to data scraping is somewhat unclear. However, it seems to imply the collection of substantial amounts of data used by artificial intelligence (AI) companies to train large language models (LLMs). These language models are responsible for powering chatbots such as OpenAI’s ChatGPT and Google’s Bard.
But what does data scraping refer to?
Source: Twitter
To put it simply, data scraping involves extracting information from the internet. For effective training, extensive amounts of authentic human conversations are necessary for large language models. But, as the quality of this data is crucial to the success of chatbots, companies working with AI often consider platforms like Twitter and Reddit as valuable sources of training data.
The thing is, these companies do not want to give this data away for free. Rather, they expect compensation for the use of their data. To ensure that organizations using their data for AI training purposes provide appropriate remuneration, these platforms have implemented specific measures. In the case of Twitter, it was restricting the access to posts.
