As AI competition continues to heat up with tech firms seeking to outdo each other, Google is working on a paid upgrade for its Bard AI.
The upgrade, known as Bard Advanced, is reportedly going to be part of Google One. It is, however, not clear if this upgrade will be accessible to all users or to select premium users.
Significance of Google One
Google One, according to the tech giant, is a “subscription plan that gives you more storage to use across Google Drive, Gmail, and Google Photos.”
When opening an account, users are granted 15GB of complimentary storage for their accounts. Users can increase the storage by upgrading to Google One. This has a storage capacity of 100GB or even higher based on the specific plan a user chooses.
Also Read: 2023 Witnessed a Surge in Major eSports Tournaments
Integration with Gemini
According to Newsbytes, “we can expect Bard Advanced to be launched sometime early this year.” The new upgrade is also expected to be available for free trials by users for a period of three months. To continue using the services, users will have to pay a subscription fee to access them.
Bard is currently powered by the Pro version in the US. A Newsbytes report further suggests that the more advanced Gemini Ultra is expected to fuel the latest upgrade, Bard Advanced.
Gemini AI was launched last December as its most powerful model, and it comes in three different sizes: Nano, Pro, and Ultra. Ultra, according to Google DeepMind CEO and co-founder Demis Hassabis, is the largest and most capable for highly complex tasks.
It uses “a combination of 57 subjects including mathematics, physics, history, medicine, law, and ethics for testing both world knowledge and problem-solving abilities,” according to Hassabis.
The tried and tested approach
The approach Google is taking with its Bard Advanced follows in the direction OpenAI has taken with its AI model. The ChatGPT maker has paid tiers for its “top-of-the-line GPT versions.”
This approach is seen as a way to offset the high costs of running large language models (LLMs). LLMs, by nature, require high energy and server infrastructure. Studies have already shown that AI firms are dealing with high energy and water consumption at their data centers due to increased demand for generative AI services.
For instance, an environmental report revealed that the amount of water consumed at Google jumped 20% in 2022 compared to the previous year, driven by the need to cool off generative AI servers.
It’s not Google alone that is dealing with the increase in consumption of water, as Microsoft reported a 35% increase in water consumption at its data centers during the same period. OpenAI had the same experience, taking up about 85,000 gallons of water during the training of its GPT-3.
The script is the same with energy consumption, which has been pointing northwards as generative AI increasingly becomes popular. Energy consumption is expected to further skyrocket with Nvidia AI chips seen taking up energy enough to power some countries.
For Google and other AI firms, introducing paid tiers could minimize some of the high costs of operating LLMs.
and then