Hello and welcome to TechScape. Johanna Bouyan and Dara Kerr are here and filled for Blake Montgomery who enjoys the beach but is likely tanned.
High-tech companies are throwing billions on AI infrastructure
High-tech companies are fighting to claim the title of having the most advanced AI in the world. The goal is to recharge your revenue and keep your investors and Wall Street happy. But developing the world's most advanced AI means spending billions of dollars on data centers and other physical infrastructure to house and enhance the supercomputers needed for AI. It also refers to the drainage of natural resources and grids in the regions surrounding data centers around the world.
Still, last week's revenue report revealed that tech companies are moving forward. Google has announced it plans to spend $85 billion in 2025 on building AI and cloud infrastructure. It's $10 billion more than we initially predicted. The company expects spending to rise again in 2026. For context, Google reported revenue of $94 billion in the second quarter of this year. CEO Sundar Pichai said that Google is in a “close supply environment” when it comes to the infrastructure needed to support AI processing and calculations. The outcome of this increased spending would still take years to come into effect, he said.
Google isn't alone. Amazon says it plans to spend $100 billion in 2025. The “majority” will utilize the AI capabilities of the cloud sector. As a comparison point, Amazon spent just under $800 billion in 2024.
“People may assume that if they can reduce the costs of any kind of technology components, they somehow lead to total spending on technology,” Amazon CEO Andy Jassy said in a revenue call in February. “We've never seen it be true.”
Meta is also increasing the amount of money they plan to spend on AI infrastructure. In June, Mark Zuckerberg said the company was planning to spend “thousands of billions” of dollars on building a network of large data centers across the United States, including what the company expects to go into operation in 2026. Originally, management was projected to spend $65 billion in 2025, but said it was adjusted to a $72bn range of 64bn and 72bn.
Meta and Amazon are reporting revenue this week.
AI is coming for artists, can you restrain it?
The artificial intelligence company fired fire to cannibalize the creative industry. Artists have seen companies use their works without permission to train their algorithms. The creative team has been reduced and fired as part of the work is being done by AI.
“That means that 95% of marketers use today's agents, strategists and creative experts, 95% are easily handled by AI, almost instantly. “No problem. “
In response, the coalition of artists launched several copyright lawsuits against top AI companies, including Openai, Meta, Microsoft, Google and Anthropic. Under the doctrine of “fair use,” businesses say copyrighted material should be available for free and without consent. Artists, including names like Sarah Silverman and Tanehishi Coates, say companies shouldn't be able to profit from their work. So far, AI companies have won.
Adobe, the software company best known for creating creative tools such as Photoshop, says it is trying to roam the boundaries of developing useful AI programs and ensuring that artists are not getting the short edge of the stick.
The company has introduced two “creator safe” tools aimed at addressing issues relating to copyright and intellectual property. One is the Firefly AI model, which says Adobe is trained only on licensed or public domain content. The other is the Adobe Content Authenticity web app. This allows photographers and other visual artists to show that they don't want to use their work to train AI. You can also add your credentials to digital creations.
Artists can apply signatures the same way that photographers sign photos and sculptors etch initials into sculptures,” said Andy Parsons, senior director at Adobe, who oversees the company's work on content reliability. We talked with Parsons about the burgeoning world of AI and what it means to creators.
***
Q: What are the biggest issues facing creators and artists with the emergence of AI and generation of AI?
I think there is one general problem. This is a concern that a variety of AI technologies compete with human ingenuity and artists of all kinds. And that applies to agents, publishers and individual creators as well.
***
Q: Is Adobe Firefly one of the ways Adobe is trying to address these issues and is making sure the creators' work doesn't tear?
Yes, absolutely. From the beginning of Adobe Firefly, we followed two guideline principles: One is to make sure Adobe Firefly is not trained with published content. They are only trained in having exclusive rights to use by Adobe and the Firefly team. This means you can't do anything specific. You can't make photos of celebrities.
The second one built in from the beginning is transparency, so we know that what comes out of the fireflies was generated by AI. This is what is called the source of content or the reliability of the content. It makes clear that something is either a photograph or something was created by an individual artist, not by AI.
After the newsletter promotion
***
Q: What does Adobe Firefly train?
This is a combination of Adobe stock and several licensed datasets. You are trained to have a clear right to use Adobe this way.
***
Q: How do high-tech companies like Adobe avoid sneaking copyrighted material into your data?set?
We have a license and explicit rights for all data contained in that dataset. We have an entire team dedicated to trust, safety and ensuring that the materials are available. We don't crave the open web. Because as soon as you do, there is a risk of potentially infringing someone else's intellectual property. Our feeling is that more training data is not always better.
***
Q: What does the future of human creativity look like now that we live in this new world with generative AI?
There is an idea of “nutrition labels” that we sometimes talk about when it comes to content credibility. Stepping into the grocery store will give you a sense of what foods are trying to serve your family, with the fundamental rights that are filled with most democratic societies. And I think the same applies to digital content. We have a fundamental right to know what it is.
UK rolls out new online safety rules after a long lead-up
Last week, the UK's internet was subject to earthquake changes. As of Friday, social media and other internet platforms must implement safety measures to protect children or face massive fines.
This is an important test of the online safety law, a groundbreaking law covering Facebook, Instagram, Tiktok, YouTube, Google and more.
Read Guardian's Guide in New Rules.
