AI Update: Google AI Content Update, AI Copyright Progress, Two New Image Generators To Try

Google's Martin Splitt Reveals How Googlebot Handles AI-Generated Content

Insights into Google's handling of AI content and quality control

In a recent webinar titled "Exploring the Art of Rendering with Google's Martin Splitt," Google's Martin Splitt shed light on how Googlebot, the search engine's web crawler, handles the increasing amount of AI-generated content on the web. The discussion provided valuable insights into Google's approach to rendering webpages and its quality control measures.

Webpage rendering is the process of assembling a webpage by downloading and combining HTML, images, CSS, and JavaScript. Googlebot follows the same process, downloading the necessary files to render webpages accurately.

During the webinar, a participant raised concerns about the impact of AI-generated content on Google's ability to render pages efficiently. The question specifically asked if rendering processes might need to be simplified due to the growing volume of AI content.

Martin Splitt responded by explaining that Google does not anticipate the need for significant changes to rendering processes in response to AI content. He emphasized that Google applies quality detection or quality control at multiple stages, and the presence of AI content does not fundamentally alter this approach.

Splitt clarified that Google's quality control measures can identify low-quality content before rendering, eliminating the need to render pages that are already deemed subpar. This detection process applies to both human-generated and AI-generated content. If a page is determined to be of low quality, Googlebot may skip rendering altogether. On the other hand, if a page appears empty or lacks clear indicators of quality, rendering may still be attempted.

While AI-generated content may increase the scale of content to be evaluated, Splitt emphasized that rendering is not the primary challenge. Instead, Google's focus remains on maintaining quality control throughout the crawling and rendering process.

Notably, Splitt did not mention specific AI detection techniques being employed by Google. However, his comments align with recent research published by Google, which revealed that detectors trained to discriminate between human and machine-written text can effectively predict the quality of webpages. This suggests that Google's quality control measures extend to AI-generated content.

The insights provided by Splitt shed light on Google's approach to handling the rise of AI-generated content. As AI continues to play a larger role in content creation, it is crucial for search engines like Google to adapt their processes to maintain the quality of search results.

As the digital landscape evolves, it is essential for SEO professionals, marketers, and content creators to stay informed about how search engines handle AI-generated content. By understanding Google's quality control measures, they can ensure their content meets the standards set by the search engine giant.

Google's Martin Splitt's remarks during the webinar provide valuable insights into how Googlebot handles AI-generated content and the importance of quality control in the crawling and rendering process. As AI becomes increasingly prevalent in content creation, it is crucial for search engine optimization strategies to align with Google's approach to maintain visibility and relevance in search results.

Can AI-generated Art Be Copyrighted? A US Judge Says Not, But It’s Just A Matter Of Time

Evelyn Waugh famously held that taking a keen interest in ecclesiastical matters was often “a prelude to insanity”. Much the same might be said about newspaper columnists taking an interest in intellectual property law. But let us take the risk. After all, you only live once – at least until Elon Musk creates an electronic clone of himself.

On Friday 18 August, a federal judge in the US rejected an attempt to copyright an artwork that had been created by an AI. The work in question is, to the untrained eye at least, no great shakes. It is called “A Recent Entrance into Paradise” and depicts a three-track railway heading into what appears to be a leafy, partly pixellated tunnel and had been “autonomously created” by a computer algorithm called the Creativity Machine.

In 2018, Stephen Thaler, CEO of a neural network firm called Imagination Engines, listed Creativity Machine as the sole creator of the artwork. The US Register of Copyright denied the application on the grounds that “the nexus between the human mind and creative expression” is a crucial element of protection.

Mr. Thaler was not amused and issued a lawsuit contesting the decision, arguing that: AI should be acknowledged “as an author where it otherwise meets authorship criteria”; and that ownership of copyright should then be vested in the machine’s owner (i.e. him); and that the register’s decision should be subjected to judicial review to clarify “whether a work generated solely by a computer falls under the protection of copyright law”.

This brings us to the district court in Washington DC and Judge Beryl A Howell, who ruled briskly that the register had not erred in denying Thaler’s application for copyright. “United States copyright law,” quoth she, “protects only works of human creation.” She did, however, concede the validity of Thaler’s claim that “copyright law has proven malleable enough to cover works created with or involving technologies developed long after traditional media of writings memorialized on paper” and went on to point out that the most recent version of the US Copyright Act allows copyright on “original works of authorship fixed in any tangible medium of expression, now known or later developed”.

In many smartphone cameras, the images are ‘post-processed’ by tiny but powerful AIsSo the law, in all its majesty, is apparently not blind to technological innovation. But, writes Judge Howell, it has always insisted that “human creativity is the sine qua non at the core of copyrightability, even as that human creativity is channeled through new tools or into new media.” Why, had not the Supreme Court itself ruled that photographs were copyrightable creations of “authors” (AKA photographers)? After all: “A camera may generate only a ‘mechanical reproduction’ of a scene, but does so only after the photographer develops a ‘mental conception’ of the photograph, which is given its final form by that photographer’s decisions.”

Quite so. But when did the Supreme Court come to this enlightened view? Er, 1884, when the court upheld the power of Congress to extend copyright protection to photography in a case involving a photograph of Oscar Wilde, no less! This is interesting because in 1884 – and indeed until comparatively recently – cameras were essentially dumb, analog machines. You pointed them at a scene, decided on the required exposure (possibly with the aid of an exposure meter), set the shutter speed and aperture, and pressed a button. The image produced by this process was chemically etched onto a glass plate or a strip of celluloid.

And now? Almost all cameras are digital and are in smartphones. You choose what you want to photograph, sure, but everything that happens from then on is done by computation. In many smartphone cameras, the images are “post-processed” by tiny but powerful AIs. (Which is why Apple has a legion of engineers working just on the iPhone camera.) The result is that it’s now rather difficult to take a “bad” photograph – one that is under or over-exposed, out of focus, or blurred by camera shake. Accordingly, most of the human “craft” of photography is taken out of your hands. And the creativity involved is boiled down to spotting an opportunity (Cartier-Bresson’s “decisive moment”, perhaps) or a scene, framing it and pressing a button. Everything else is done by AI.

Still, at the moment it meets the “human involvement” criterion of the 1884 judgment and of contemporary copyright legislation. But my hunch is that its days are numbered. Indeed, even Judge Howell seems to agree. “We are approaching new frontiers in copyright,” she writes, “as artists put AI in their toolbox to be used in the generation of new visual and other artistic works. The increased attenuation of human creativity from the actual generation of the final work will prompt challenging questions regarding how much human input is necessary to qualify the user of an AI system as an ‘author’ of a generated work.”

She’s right. And – who knows? – If he lives long enough, Mr Thaler may even get his copyright.

New Image Generator That’s Text-Friendly

Check out this review of IDEOGRAM…

New Way To Create Consistent Characters

Free alternative to Midjourney built on top of Stable Diffusion…