The New ‘Face’ of AI

Harvard dropouts seem to do well for themselves: Bill Gates, Matt Damon, and Facebook CEO Mark Zuckerberg. In terms of talking about AI, Facebook has been in the news lately as Zuckerberg revealed at the Townhall Q&A how AI and VR will be a big part of our tech future. The tech billionaire has grand designs to one day integrate AI into his social media empire, which he hopes will further connect people.

facebook-logo

As the largest tech enterprise in the world, Facebook may be the precursor to what lays ahead for the tech industry. The company has been working on improving artificial intelligence, even planning to establish a special laboratory in France dedicated to research and studies involving the technology.

Zuckerberg admitted that in order for AI technology to work, it must be smarter than the lay human. AI must be able to understand text posts, recognise people and events from photos and videos, and be mentally able to process them.  Zuckerberg thinks this can happen within 10 short years; and he may not be that off with his estimate, as even Google has positioned itself in the AI game.

Facebook has seen success in their artificial intelligence research, as Facebook Inc. has recently revealed a series of photographs produced by an AI living in the Menlo Park laboratories. According to the study, the machine was able to recreate images that were identical to the original picture. About 40% of the humans polled believed that the AI replicated images were real.

But Facebook Inc. has gone beyond testing the ability for machines to copy photos, they have also employed the AI technology within their website’s algorithms. The photo-tagging feature is capable of identifying users by simply analysing the photographs uploaded throughout the site. Facebook will even improve facial recognition by being able to distinguish a person’s face even if the image is partially concealed.

If Zuckerberg’s estimates are right, the tech landscape will be rampant with AI tech in 10 years. It will be interesting to see if AI ‘face-tagging’ technology will recognize what the ‘face’ of Silicon Valley was a decade previous.

Advertisements

U.S. Cyber Attack Reveals Continued Vulnerabilities

Avi Yashchin Cyber AttacksAs media outlets continue to learn more about the unprecedented scale of the recent cyber hacking on United States government employees, fears continue to mount in both the public and private sector. With an estimated 18 million government employees compromised in the attack, it’s easy to see why fears are mounting. Arms industry executives have expressed frustration as to how such a large scale cyber attack could have happened in the first place—especially since many of those hacked work on high security arms projects.

As China continues to deny culpability for the intricate spearfishing hack, the larger issue about government employees’ security and protection remains. The United States Office of Personnel Management (OPM) faces mounting concerns, as well as ongoing congressional hearings, about its security measures taken to prevent such attacks. Aside from access to Social Security numbers, data obtained by the hackers could be used to purloin highly sensitive government weapons program profiles, passwords and files. This chink in the United States weapon industry’s armor leaves U.S. weapons networks vulnerable to future attacks.

The real question is: could this have been averted, and how? Dave Wajsgras, head of Raytheon Co’s Intelligence, Information and Services business, told Reuters that he had previously raised alarms about pre-existing private and government breaches, and was distressed how easily OPM was hacked.

“There is a tsunami of threats that exist in the cyber domain today. It’s something that we all collectively need to take much more seriously,” he added, suggesting more spending for proactive bolstering security in government and the private sector.

Even after network breaches of the U.S. Navy and U.S. Marine Corps servers last year, the OPM’s network was reportedly not encrypted. These continued hacks raise increasing concerns about the level of (in)security maintained by the agency.

After a weapons tester reported network vulnerabilities in weapons programs, the U.S. Senate Armed Services Committee has proposed a $200 million Pentagon’s fiscal budget increase next year. But whether injecting more funds into a failing cyber security system will deter an indefatigable foe will suffice is yet to be determined.