[ad_1]
It has already been known that artificial intelligence (AI) models are trained on data sets that allow them to analyze and respond to texts in several contexts as well as languages. For instance, ChatGPT has been trained on a massive text dataset that is available in the public domain. On the other hand, DarKBERT is an LLM that has been trained on a vast dataset of dark web pages, assimilating information from places such as hacker forums, scamming websites, and other criminal internet sources. Since AI tools’ hunger for data is insatiable, everything posted online by anyone, is fair game. Making that clear was Google, which has updated its privacy policy, and everything that you post online, could now be used to train its AI tools and models.
New privacy policies
Google announced changes to its privacy policies on its website. It states, “Google uses the information to improve our services and to develop new products, features, and technologies that benefit our users and the public. For example, we use publicly available information to help train Google’s AI models and build products and features like Google Translate, Bard, and Cloud AI capabilities.”
One look at Google’s privacy policy history details the changes that Google has made. Previously, Google stated that your data might be used for “language models”, however, it has now been replaced with “AI models”. Moreover, the policy only mentioned Google Translate earlier but Cloud AI and Google Bard have now been included.
While most companies’ privacy policies include the right to use any data posted on their platforms, Google now reserves the right to gather and use data posted on the web as a whole that will be used to develop its services and train its AI models.
How do AI models get their data?
Generative AI models like ChatGPT use the whole internet to source their data through a process called Web scraping. It extracts a valuable amount of data from online sources and then provides a sentiment analysis of the same to the user. While web scraping can be useful for analytical research purposes, it can also violate the terms of service of a website that prohibits web scraping.
To counter the extreme levels of data scraping and system manipulation, Elon Musk recently restricted Twitter accounts to a limited number of readings per day. Moreover, Twitter also restricted browsing access for users without accounts. Know more about this here: Elon Musk changes Twitter forever, slaps limits on number of tweets you can read.
[ad_2]
Source link