Like Microsoft, Google claims it will still run all of this using clean energy by 2030. However, the current emissions data do not indicate that this is possible. The need for energy will not go down, and these companies will not walk away from the trillions of dollars invested in AI already.
This is happening while we watch the earliest recorded Category 5 hurricane because of the extremely warm water temperatures in the Atlantic.
Sharlyn chalks this up to AI still learning, but I don’t believe that’s correct. I don’t think Generative AI tools can come up with anything new. That’s not the way it works. Large language models are developed by ingesting and analyzing tons of existing data. Based on that data, they then mathematically conjure up the most likely response to your prompt. How could it come with a genuinely original icebreaker?
In 2020, Microsoft promised to be at negative carbon emissions as a company in ten years. Instead, they have increased emissions by almost 30% since then. They promise they’ll be there in 2030, but that seems like wishful thinking.
This story seems inevitable to me. As the tech world pushes us increasingly into asking an AI chatbot for information instead of looking it up ourselves, and the AI model has incorrect information, where do you get it corrected? When OpenAI was asked to correct or remove this misinformation, they said it was ‘technically impossible’…
When management harms the mental health of our employees, we typically respond by offering them yoga or meditation spaces or maybe a lunchtime session on stress management. We never look at the system. We offer them ways to better cope with the broken system, but we never take responsibility for what the workplace is doing to their mental health.
Rolling out an AI model might turn into much more of an Information Governance exercise than you’ve bargained for. Maybe that’s a good thing, though. Maybe it’s time to get rid of all that junk, not just because of the junk itself but because it’ll make your AI rollout more successful.