Beyond Tomorrow: AI Ads, Google's Carbon Emissions Balloon, Apple Joins OpenAI Board — Weekly AI Roundup

Everyone and their mother is excited about artificial intelligence thanks to how fast the tech is evolving. But what about the climate cost?

Big Tech has made promises to become net-zero with regards to carbon emissions, but with how much computing power is required to train AIs, is it a realistic goal? Google and Microsoft Inc.'s sustainability reports paint a bleak picture in that regard.

In other news, Apple Inc. is reportedly going to be getting an observer role on OpenAI's board. The $3.4-trillion company won't get to make decisions regarding the AI venture's direction. But the board observer role will definitely give it insight into how OpenAI is run.

Finally, children's toys retail chain, Toys 'R' Us, ran an AI-generated advertorial. Predictably, the ad faced backlash. However, the company's president told NBC news in an interview that the campaign was a "success". Now, Motorolla has stepped up with a 30-second ad campaign which, honestly, looks pretty slick. These kinds of ads have larger implications which need to be discussed before it's too late.

Here is some AI news from the past week that's worth a read.

Every year Google releases their environmental report, in which they describe what they’re doing to reach their net-zero emissions goal. This year’s report doesn’t exactly paint a rosy picture of the Silicon Valley giant’s lofty aim.

To put it into context: Google plans to achieve net-zero emissions across all its operations by 2030, with 2019 as the base year. Some would say that is a reasonable goal.

However, the company reported that it had seen a 13% year-on-year increase in green house gas emissions in 2023. Just last year, the tech giant's operations produced pollution equivalent to 14.3 metric tons of carbon dioxide pollution. And it gets worse.

Google’s green house gas emissions have grown 48% since 2019.

The company has attributed the increases to “data centre energy consumption and supply chain emissions.” In non-corporate speak, Google’s AI push isn’t doing any favours for the government. Data centres use a massive amount of electricity and generate a ton of emissions. Training AI just makes those costs go up even more. The Verge has a great breakdown of how much elecricity does AI consume.

The company itself has admitted this in the report, saying:

"As we further integrate AI into our products, reducing emissions may be challenging due to increasing energy demands from the greater intensity of AI compute, and the emissions associated with the expected increases in our technical infrastructure investment."

Without condoning Google, these sort of emissions are inevitable, given how the tech industry as a whole is trying to reach the as-yet undefined finishline for AI.

To be fair, it's not like the company isn't trying either. In their report they've outline several strategies they've either undertaken already, or that are in process to make their business carbon energy free.

Some of these include optimising their data centres to be more energy efficient, as well as finding sources of carbon-free energy across the Americas and the Asia Pacific region.

Google isn't alone in this, Microsoft is also struggling to keep pace with its climate goals. The Bill Gates-founded company has its own climate goals, with its own annual sustainability report. With a 2020 baseline, Microsoft has found that its emissions since that year are up 29.1%.

Apple and OpenAI are getting real cosy. The tech giant is likely to get an observer role on the AI-focused company's board, according to Bloomberg's Mark Gurman. The position on the board is a part of the two firms' agreement to integrate ChatGPT into the Apple ecosystem.

So, who's going to be taking on the "board observer" role? According to Gurman, it is Apple's head of the App Store, Phil Schiller. So far, he hasn't attended any board meetings but that might change.

As a board observer, Schiller will be allowed to attend OpenAI's board meetings, but beyond that, he can't vote or exercise powers another director might have. What it does allow him to do is to watch how the company is run and report back to his bosses.

That being said, OpenAI's decision to let a member from Apple onto its board is a sign. The company might be trying to assert some sort of independence from its largest backer, Microsoft. It's definitely a decision that will likely raise some eyebrows at the offices of the creators of the Windows operating system.

But strange times make for even stranger bed fellows, right?

Last week, Meta Platforms Inc. and Apple were reportedly in talks to include the Llama large language models into the latter's ecosystem. It has so much potential to be a strange pairing. Doubly so, considering that Apple has quite openly criticised the Zuckerberg-led company's privacy practices.

The fact that Apple is still taking meetings with Meta as well as Anthropic, despite the OpenAI agreement, shows that the company doesn't want to be beholden to anyone.

Fortune's David Meyer did a neat little breakdown of what the board role means for Apple, Microsoft and OpenAI. Give it a read.

Remember when we first started seeing examples of images and videos generated by AI? We were all confident that the technology won't be replacing people who produce “creative” work.

If Motorolla's advertisement for the Razr50 is anything to go by, it might be time for us 'creative types' to redefine our relationship with the technology. The 30-second 'Styled With Moto' campaign used at least nine genAI video generation tools including Midjourney and Sora, according to a report from Marketing Dive.

The clip took four months to make—from research to production—given that the company's agency, France-based Heaven, had to sift through thousands of AI-generated images to find the looks they wanted for the campaign. One can't help but wonder if this would have been faster to do with a human team?

The video itself is pretty slick. If you're not outright paying attention and looking for the telltale signs of AI-generated content, there's a high chance you miss it.

Most importantly. though, as mentioned earlier, what happens to the people who work on these kinds of shoots? That list includes producers, camera and sound operators, directors, graphic artists, and countless others. What happens to them? Are creative industries even ready for this kind of technology? Have they found any way to find a place for themselves alongside generative AI?

There's a lot of conversation about these topics that desperately need to be had, which still seem to be in their nascent stages in India. While we still haven't reached a point where genAI can make me laugh with a very pointedly political comic, how far are we really from something like that?

Beyond Tomorrow is a weekly newsletter sent to your inbox every Saturday to give you a roundup of everything AI in the last week.