Human intellect falling behind AI? Even major announcements go unnoticed
A new machine-learning API that takes video content analysis to the next level was launched last week at Google Next. Artificial intelligence is advancing by leaps and bounds, but the algorithms hidden at the heart of the data center have not captured the imagination of the general public.
Google’s whole AI product family receives surprisingly little press, even though it will be one of the first to make machine learning available to practically everybody. A fine example of this lack of attention was last week’s Google Next, the equivalent of Apple’s massive annual launch events.
In addition to new partners, the event saw the announcement of a swarm of new products and add-ons, product upgrades and the latest achievements of Google Brain and TensorFlow.
“It boggles the mind that there isn’t more coverage”, says Jerry Jalava, Senior System Architect and Google Developer Expert at Qvik. “A lack of understanding of the current potential of machine learning is a big obstacle to progress in many organizations.”
One of the most intriguing announcements was Video Intelligence API. This machine-learning API enables the multi-stage content analysis of video materials with a single API call. Qvik immediately adopted the API in its customer projects.
The development of video analysis based on machine learning is a big deal. There is so much video material on the internet that humans could never go through all that data without help. Every minute, 300 hours of video is uploaded to Youtube alone. That’s a lot to tackle, even with an AI.
GOOGLE ON A ROLL: ROUGHLY ONE HUNDRED OTHER ANNOUNCEMENTS MADE AT NEXT
Google’s cloud service and machine learning offering is expanding at an amazing pace right now. One hundred announcements is a lot to take in, so here are the seven we thought the most interesting in addition to Video Intelligence. Click on the links to read more.
Our picks are of particular interest to infrastructure and developer teams, data analysts, companies that process confidential materials and everyone who handles data keys.
GRAPHICS PROCESSING UNIT (GPU)
Google Cloud received the long-awaited cloud-based graphics processing units (GPU), which make scientific calculations and machine learning considerably more efficient.
KAGGLE JOINS GOOGLE CLOUD
The Kaggle community includes more than 800,000 data scientists, with whom Google hopes to advance its mission of democratizing artificial intelligence.
IDENTITY-AWARE PROXY (IAP)
This IAP provides companies with fast and secure implementation services. With it, they can offer access to their applications without having to code through the authentication level.
DATA LOSS PREVENTION (DLP)
DLP identifies data that requires encryption, such as credit card numbers and personal data, from images and text. The confidential material gets put through an API, which redacts the data to a safe form before returning it.
KEY MANAGEMENT SERVICE (KMS)
KMS maintains a database of sensitive keys, ensuring its technical integrity and preventing leaks.
GOOGLE-DESIGNED SECURITY CHIP
Google has always been particular about data security. Now, the company has announced the Titan product, which will take data security to the hardware level. The product consists of a chip the size of an earring that is physically installed into devices, guaranteeing more secure authentication.
CLOUD DATAPREP (PRIVATE BETA)
Cloud Dataprep, currently still in beta, will offer powerful assistance to data analysts. The tool takes the visual exploration and preparation of data to a new level.
Last week, more than ten thousand technology professionals travelled to San Francisco for the Google Cloud Next event, including Qvik’s Senior System Architect, Google Developer Expert Jerry Jalava and CEO Lari Tuominen.