• Nov 5, 2020

Artificial Intelligence by Google

Original article 5  Nov 2020 - Updated Aug 2024

Artificial intelligence (AI) is enhancing our lives and augmenting our abilities.

AI for Hearing Speech

It is not until you have a hearing problem, that you fully understand the issue (difficulty) of hearing in background noise.

So when Google recently announced a technology to improve speech understanding in video, many people with hearing problems recognised the similar technologies used to achieve this.

"By training the model on a large-scale collection of online videos, we are able to capture correlations between speech and visual signals such as mouth movements and facial expressions, which can then be used to separate the speech of one person in a video from another, or to separate speech from background sounds."

It appears Google has developed technologies that 'lip read' or 'utilise facial cues'. Also looking at Voice Pitch and Age is related to how hearing aids function. Signal-to distortion ratio (SDR), in dB for different spoken languages was also compared.

"As our method utilizes facial cues and mouth movements to isolate the speech, we tested whether facial hair (e.g., a moustache, beard) may obstruct those visual cues and affect the method’s performance. Our evaluations show that the quality of speech enhancement is maintained well also in the presence of facial hair."

Hearing Aids

Similar technologies are used to enhance speech in hearing aids. The largest improvement in hearing aids is in separating speech from noise.

Phonak in 2024 have introduced a hearing aid with on board AI to reduce background noise with its own dedicated chip.

AI for Google Search

Spelling in search queries

Approximately 1 in 10 queries typed into Google are spelt incorrectly. AI allows Google to make sense of wrongly spelt words, and usually calculate what is meant, to then deliver the correct search results.

We adapt and get used to this. We also change our queries to give the search engine the correct context as quickly as we can.

SEO in a blog

As a Gold Coast based SEO we recognise the value of Google appreciating a well researched article.
Otherwise SEO's would get away with just putting a bunch of keywords into a blog. It would not have to make sense, be grammatically correct or reach a logical conclusion.

Many articles are not actually written to be read, they are written to fool search engines into increasing the ranking for a site. Google AI differentiates well written articles and ranks them higher.

Google's SpamBrain

SpamBrain is Google's AI detection mechanism for spammers and low-quality websites.

"Even if you are not a spammer Google can recognise low quality content"

"Google also highly discourages using techniques commonly associated with spam or low-quality content. These include manipulative keywords, cloaking, hidden text, and too many ads."

The algorithms "learn" from your website, and it improves over time. If it finds that your site is still spammy, it'll continue to punish you.

Google used this AI based spam-prevention system for the first time December 2022.

Image and Video Recognition

Google uses AI to understand the content of images and video. Say there are 2 articles with similar content, the article with unique and relevant images and or video will rank higher.

Given the abundance of available text via AI systems like Chat GPT, Google will be integrating systems to identify purely copied content. One obvious technique would be to identify quality formatting of articles and relevant images and videos.

The Downside of AI

Adrienne Williams argues that 'these' [AI] systems are fuelled by millions of underpaid workers around the world, performing repetitive tasks under precarious labour conditions.

"They are striving to create autonomous systems that can one day perform all of the tasks that people can do and more, without the required salaries, benefits or other costs associated with employing humans. While this corporate executives’ utopia is far from reality, the march to attempt its realization has created a global underclass, performing what anthropologist Mary L. Gray and computational social scientist Siddharth Suri call ghost work: the downplayed human labour driving 'AI'."