Google's decision to collect public data to train AI could have huge ramifications

Google's recently modified privacy policy sheds light on the search giant's growing AI ambitions. According to the updated privacy policy, Google reserves the right to scrape just about anything posted publicly online and use it to improve its AI tools like Google Bard.

Back in May, developer DataChazGPT indicated that Google's new search engine called Project Magi will compete with Microsoft's Bing Chat and OpenAI's ChatGPT chatbots. The word on the street is that Project Magi will make search more visual, personal, and human with the help of AI.

Now, the folks at Gizmodo have spotted Google's updated policy, which states the company is also planning to use "publicly available information to help train Google's AI models and build products and features like Google Translate, Bard, and Cloud AI capabilities."

To recap, the previous policy wording suggested that Google simply uses the information to build "language models" for its Translate service. However, the updated policy now includes Google Bard, as well as the term "Cloud AI." So, what does that mean?

What does Google's updated privacy policy imply?

Anything you post using a Google product like YouTube, Gmail, or Search will be saved and used by the American tech giant in some way or another. Notably, the wording explains that Google will use publicly available information to train its AI.

In other words, Google will use anything posted on the internet. This is part of the tech giant's attempt to improve its AI tools. In line with this, Google recently introduced a new Bard feature that allows users to visualise their search results. However, it is safe to say that Google's latest AI move is likely to have a more prominent impact.

The report points out that Google's Search platform, which was built on PR (PageRank), can effectively index the web. SearchEngine Journal describes website indexing as one of the initial steps in the process of how web pages are ranked and displayed as search engine results.

Google announced nine notable AI features at the recently concluded Google I/O. For instance, the company introduced Duet AI in Google Workspace. However, many organisations are worried about where the LLMs (large language models) that power these projects are getting their information.

To make things worse for AI companies, the EU is prepping to impose strict regulations on AI products. In fact, Google had to stop the rollout of its Bard AI in Europe last month. On top of that, Reddit, Twitter, and other similar platforms have refused to offer free access to their APIs.

Apparently, these APIs come in handy for downloading huge backlogs of posts and potentially collecting the content to train AI. Similarly, Google's biggest rival in the AI market, ChatGPT had to introduce data controls for users worldwide citing a complaint from Italy's data protection authorities.

The negative impacts of AI

These restrictions enable users to stop OpenAI from training ChatGPT on the data they provide. Still, it is not as big as Google's latest move to include the entire public internet to build and improve its AI tool. Meanwhile, a CNN report suggests a considerable number of layoffs occurring in the tech sector are linked to AI.

Several companies have paused hiring while terminating their employees as they try to figure out which roles can be taken over by AI. For example, IBM CEO Arvind Krishna told Bloomberg that the American technology corporation was planning to freeze its hiring process while it tries to understand where AI's role will be more useful.

Notably, the universities in the UK have come up with guiding principles for generative AI. The goal is to spread awareness of these technologies. In the meantime, universities are struggling to adapt teaching methods due to the skyrocketing popularity of AI, according to a report by The Guardian.