You may have noticed the recent attempts by Google to provide AI-generated answers to your queries. You might not immediately understand why AI is such a prevalent thing in arbitrary technology like search engines, but we assure you that there is a good reason for it—a better user experience—even if it does come with mixed results.
Google Just Wants to Help Its Users
There’s a reason why Google has been most people’s preferred search engine, and it’s because it provides a great user experience and the best search results. Google is remarkably helpful for many tasks, even those that you might not expect. It can give you the approximate calories of a particular food, as well as its entire nutrition label and macronutrient breakdown, or it can give you the answer to a math equation.
It’s almost like Google is trying to cut itself out as the middle-man and make itself the go-to resource for super simple queries.
Of course, not all questions are simple. Google is making strides toward tackling some of the more complex ones with its new AI product, Gemini… often with hilarious results.
Gemini’s Advice is… Not Good
What seems like a logical answer to computers is far from logical to humans. This is a truth seen in the way Gemini used to answer questions in its earlier days.
For example, back in May, if you asked Google how to get cheese to stick to pizza, the AI would suggest using non-toxic glue.
I don’t know about you, but I don’t want glue in my pizza, toxic or not. So why would Gemini offer this solution?
All it took was a single nonsensical post on Reddit to make AI short-circuit its logic cylinders and give up a half-baked response.
While Google has been trying to resolve these issues, it’s experienced mixed results. You’d think with this level of technology, it would be able to learn more from us, but alas, this is not necessarily the case.
And it sets a dangerous precedent.
Google is a great resource for people, so much to the point where they depend on it to be correct. In certain cases, it can be fatal, like in the case of toxic or non-toxic foods. The Associated Press did a test that confirmed this with mushrooms in the forest, and it found that AI, while somewhat accurate, could also be dangerous.
Worse yet, AI cannot tell the truth from lies or conspiracy theories, which is not helpful in combating misinformation. It also cannot determine what is good advice or a joke on the Internet. Using generative AI requires significant responsibility and understanding on the user’s part of what AI can and cannot do.
AI Responses Eat Up Energy
Google Search has been optimized for efficiency, and with the way that Google builds data centers, its redundancy and content delivery systems, and other solutions keep Google’s costs and energy usage lower than you might think. Despite this, Google is massive, and in 2011, Google’s entire infrastructure consumed about the same electricity as 200,000 homes—about a quarter of the output of a nuclear power plant. Since then, the number has only increased. In 2019, Google used more electricity than some small countries, and if Google were its own nation, it would be in the top 90 of electric usage.
Even with Google making the most of its infrastructure, the number of services and solutions it offers continues to climb, and more people become reliant on it day by day.
This includes AI.
A decade ago, a Google search would use about 0.3Wh of energy—enough to power a low-energy 10 watt bulb for about a minute and a half. An AI response, however, uses enough power to run a 22-inch LED TV.
Google receives 99,000 search queries each second, or an average of 8.5 billion searches per day. With AI showing up periodically in the results, you can imagine the amount of electricity that’s consumed.
AI is Bad for the Environment
Most of the AI processing that happens these days is done in data centers owned by Google, Microsoft, OpenAI, Amazon, and other large entities. It’s not done on your smartphone or workstation.
But these data centers are growing larger and larger, overtaking electrical grids and becoming more energy-demanding. The resources to power these data centers—the fuel, the water to cool the servers, the air conditioning units to keep the environment tech-friendly, all of it. It’s shocking what goes into powering these massive systems.
Is it worth it for bad advice on the Internet?
What are your thoughts on using AI effectively, ethically, and efficiently? We’d love to help you improve your AI practices. To learn more, call us at 781-679-0172.
