Apple just put AI in millions of people’s pockets. The company is rolling out what it calls Apple Intelligence This week, iPhone, iPad and Mac users are getting some basic text generation and image editing features Those who choose. I’ve been testing these tools through the developer beta version of the software for a few months now, and they’re pretty mediocre. But this is only the beginning.
Generative AI, once a parlor trick for the tech-obsessed, is fast becoming the main event for major software releases. As Apple pushes its version of technology, Google Building AI into its Android operating system And forces everyone to see the AI overview at the top of virtually every Google search OpenAI And meta The startup is building its own AI-powered search engine Confusion Already have one. Microsoft And anthropological Recently announced are new, super-powerful AI agents that can perform complex tasks just like humans. (Disclosure: Vox Media is one of several publishers that have signed a partnership agreement with OpenAI. Our reporting is editorially independent.)
While some companies’ generative AI products have been in the wild for more than a year, the arrival of Apple Intelligence marks a turning point for the technology’s mainstream. Apple Intelligence is only available on the latest Apple devices, but on more than half of the phones in the US iPhone is. People can tap into millions more new technologies as they upgrade
If you’re not already using AI, you probably will be soon – whether you like it or not.
“We’re getting AI, especially generative AI, shoved down our throats with little or no transparency, and frankly, opt-out mechanisms are either non-existent or complicated,” said Sasha Lucioni, AI researcher and climate lead for AI and machine learning at Hugging Face. Platform for sharing tools
If this fills you with dread, it’s understandable. Maybe you feel bad about participating in a race to build a super-intelligent AI that no one wanted. You may feel complicated to use AI models Trained on copyrighted material Without paying the manufacturers. You probably just feel general bad about the flood The AI slop that’s ruining the internet Even if you don’t personally make slop.
Then there are all the climate consequences. AI, in its many shapes and forms, requires a lot of energy and water to function. A a lot. It can make you feel guilty right away about using AI.
AI’s big power hunger
In terms of power, Apple Intelligence has the potential to be far more crime-free than other major AI alternatives. Apple says it puts the processing for certain AI features, such as GenMoji and Image Playground, entirely on your device. This means less reliance on energy-intensive data centers.
We don’t know exactly how much power AI consumes in these data centers. Using data from a recent Microsoft Research study, University of California Riverside engineering professor Shaoli Ren came up with this: Asking ChatGPT to write two 200-word emails would use about the same amount of energy as a Tesla Model 3. run a mile Because they generate so much heat, the processors that generated the emails will also need it About four and a half liter bottles of water to cool down
If you scale up, the consequences of such power trends become more apparent. The amount of electricity used by data centers, where AI processing mainly takes place 160 percent growth forecast By the end of the decade, and carbon dioxide emissions could more than double as a result, according to Goldman Sachs. Meanwhile, the amount of water needed will also increase, so much so that by 2027, AI’s thirst That could equal half of the UK’s annual water withdrawals.
These are all estimates based on limited data because tech companies that build AI systems, including Apple, Google, Microsoft and OpenAI, don’t share exactly how much energy or water their models use.
“We’re just looking at the black box because we have no idea about the energy consumption of interacting with large language models,” Ren explained. He compared the situation to searching for flights on Google and being able to See carbon emissions for each leg. “But when it comes to these large language models, there’s absolutely nothing, zero, no information.”
A lack of transparency about AI’s energy needs runs counter to these tech companies’ commitment to sustainability. There are good reasons to believe that AI Those who break promises are directly led.
As data center energy use increases, Google has seen its greenhouse gas emissions rise 48 percent increase from 2019 to 2023, despite its pledge to cut emissions by 50 percent from 2019 levels By 2030. Company And does not claim to be carbon neutral. Likewise Microsoft saw a 29 percent jump in emissions From 2020 to 2023. Microsoft has pledged to be carbon negative by 2030, it is Now struggling openly Keeping up with AI innovation and ways to make it happen.
What AI dealers aren’t telling us
This is what an arms race looks like. It should be noted here that all energy consumption started to increase during OpenAI Knocked the socks off the world With the surprise release of ChatGPT in November 2022. Chatbot Has become the fastest growing app everCapturing over a hundred million users In two months And AI started the gold rush in Silicon Valley. now, 40 percent of all venture capital Cloud computing goes to generative AI companies. OpenAI itself announced a $6.6 billion funding round in early October — The largest venture capital round of all time — giving it a $157 billion valuation.
With such a staggering amount of money at play, it’s perhaps no surprise that energy efficiency takes a back seat to innovation. Companies like OpenAI want models that power their AI technology so they can get better and outperform competitors. And the bigger the model, the higher the energy demand — at least for now. Over time, performance is likely to become more efficient thanks to advances in chip technology, data center cooling, and engineering.
Josh Parker, head of sustainability at chipmaker Nvidia, told me, “Because innovation happened so quickly when ChatGPT burst onto the scene, you’d initially expect efficiency to be at a minimum.
Still, the most energy-intensive products are now what are pushing companies like OpenAI, Google, and Meta the most These include real-time chatbots, voice assistants and search engines. This feature Larger models are listed and require more advanced chips Working at the same time to reduce delay or lag. Simply put, they have to do many difficult math problems simultaneously and very quickly. That’s why it takes as much electricity to run a Tesla.
Apple, however, seems to be presenting itself as an exception. As part of its commitment to protecting user privacy, the company says it handles as many Apple Intelligence tasks as possible on your device without sending queries to data centers. This means when you opt in to Apple Intelligence, you download a small generative AI model that can handle fairly simple tasks on your phone. Your iPhone battery, unlike a grid-connected cloud data center, has a limited amount of power, forcing Apple Intelligence to handle these tasks with some efficiency. Perhaps on-device AI is the crime-free version of the future.
The problem, of course, is that we don’t know exactly how Apple Intelligence works. We don’t know what tasks are handled on the device, which are sent to the power-hungry Apple servers, or how much power it requires. I asked Apple about this, but the company did not provide specifics. Then again, not providing specifics is a bit of a theme when it comes to big tech companies explaining their AI offerings.
So again, it’s understandable if you feel fear or guilt about AI in your life. It is abundantly clear that this technology, in its current state, consumes enormous and increasing energy, contributes to greenhouse gas emissions, and worsens human-caused climate change. It’s also true that you may not have a choice, as large tech companies make generative AI more fundamental to their products. You can opt out of Apple Intelligence or never opt in But you’ll find it harder, if not impossible, to opt out of AI products from Google, Meta and Microsoft (If you want to try, here A helpful guide.)
“I don’t think there’s any reason to feel guilty,” Lucioni said. “But I think there’s a reason — like climate change in general — to ask for more information, to ask for accountability on the part of the companies that are selling us these things.”
If AI solves all our problems or destroys all or both of us, it would be nice to know the details. We could query ChatGPT, but that would be a huge waste of energy.
A version of this story also appeared in the Vox Technology Newsletter.Sign up hereSo you don’t miss the next one!
