Artificial Intelligence: Creepy or Helpful?

Posted by ABS Tech on July 11, 2018 with 0 Comment

Artificial intelligence and machine learning has become a hot button topic in Technology. Apple brought us Siri, Amazon brought us Alexa, and Google brought us Google Assistant. But AI capability and purpose goes far beyond your common answer to “what’s the weather like?” and “set a reminder when I arrive at home”. AI is ready to help businesses trace food from farm to store, give a guided tour through New York City, research the cure to cancer, and to secure our networks. However, AI is something many of us fear. Voices throughout the technology community have expressed their warnings of further research on AI. Stephen Hawking said Artificial Intelligence will be the end of man”. Elon Musk stated, “We are rapidly headed towards digital super intelligence that far exceeds any human”. Should we mind these warnings?

Here’s some history.

In 2011, IBM’s Watson won the game of Jeopardy against 2 top human challengers.

Not only did Watson beat top challenger’s in Jeopardy, Watson brought deep questions about our limitations as humans and the capabilities of machines.

In 2017 Facebook created two AI chatbots, which were trained to have certain likes and pretend to express some sort of grief when they lost what they wanted. In an experiment, the AI’s were challenged to speak to one another and negotiate a trade of balls and books. The results of the conversation are seen below (One AI was called Bob, the other Alice):

Bob: i can i i everything else . . . . . . . . . . . . . .

Alice: balls have zero to me to me to me to me to me to me to me to me to

Bob: you i everything else . . . . . . . . . . . . . .

Alice: balls have a ball to me to me to me to me to me to me to me

Bob: i i can i i i everything else . . . . . . . . . . . . . .

Alice: balls have a ball to me to me to me to me to me to me to me

Bob: i . . . . . . . . . . . . . . . . . . .

Alice: balls have zero to me to me to me to me to me to me to me to me to

Bob: you i i i i i everything else . . . . . . . . . . . . . .

Alice: balls have 0 to me to me to me to me to me to me to me to me to

Bob: you i i i everything else . . . . . . . . . . . . . .

Alice: balls have zero to me to me to me to me to me to me to me to me to

What may seem like a useless conversation was actually the two bots communicating in their own language. The creators shut down the bots as they had no use to humans.

So AI is pretty smart right? But can modern AI beat the Turing test? The Turing test was a test developed by computer scientist Alan Turing (known for breaking the enigma encryption algorithm during World War II) used to distinguish whether a computer is artificially intelligent, it requires that a human being should be unable to distinguish the machine from another human being by using the replies to questions put to both.

Can you distinguish the Google Duplex AI from a human caller?

OK, Maybe this is all a little creepy.

So maybe we cease to use technology altogether until every programmer vows to never create AI again; their creepy computer voices won’t be able to laugh at us in binary as we post silly human memes across their domain. Do we smash every smartphone? Every smart speaker? Every Internet connected device? Do we stop progress? Was Elon Musk right? Those are all questions for a philosophy course. For now, there are many benefits we can gain from utilizing AI.

Apple Siri, Amazon Alexa, Google Assistant, and Microsoft Cortana serve as assistant AI’s to help keep you organized, and keep your technology working for you in very personal ways. Read more about them in my Tech Talk on smart things and privacy

Barracuda’s Sentinel AI was build to detect phishing scams so we don’t have to figure it out. It learns the patterns to the emails we send and receive and is able to determine whether a message is a spoof message to or from you.

There are many ways phishing scams can come through. And utilizing AI to assist with detecting it is the best way to avoid falling victim to it.

In conclusion, Modern AI is smart, and keeps getting smarter. The best way to stay ahead of the machines is to utilize them to your benefit.

Sources:

https://www.bbc.com/news/technology-30290540

https://gizmodo.com/nvidia-taught-an-ai-to-flawlessly-erase-watermarks-from-1827474196

https://www.forbes.com/sites/bernardmarr/2018/07/10/the-economics-of-artificial-intelligence-how-cheaper-predictions-will-change-the-world/#50fdd7655a0d

https://www.ibm.com/blockchain?lnk=ushpv18c10&lnk2=learn

https://www.independent.co.uk/life-style/gadgets-and-tech/news/facebook-artificial-intelligence-ai-chatbot-new-language-research-openai-google-a7869706.html

https://ai.google/research/

https://www.washingtonpost.com/news/innovations/wp/2018/04/06/elon-musks-nightmarish-warning-ai-could-become-an-immortal-dictator-from-which-we-would-never-escape/?utm_term=.1a8fc6b6310d

Privacy Matters

Posted by ABS Tech on February 13, 2018 with 0 Comment

Privacy. It’s a minimum requested to service providers when on the web. But what does privacy compromise on the internet? There are a few parts to internet privacy you may have heard of before: Cookies, Private browsing (aka incognito in Google Chrome), and VPN connections. But what does this all mean? Does it really protect you? And what is it protecting you from?

Let’s start with what information is collected

Most information collected is used to send ads that are more relevant to you that you will be more likely to click on. This information is extremely valuable and the relevance of these ads and whether they are clicked on has a direct impact on company profits. This of course is unmalicious in nature. The question is what information are they using and whether you want that information collected.

So, let’s talk about the methods used to obtain this information. Some methods are cookies, and fingerprints.

Cookies:

A cookie is a small file that a website downloads to your web browser that collects data about you. Some cookies are necessary, such as an authentication cookie that tracks whether you login or log out of their site. Other cookies are unnecessary such as third-party tracking cookies, which can be issued by a site other than the site you are visiting.

Fingerprints:

Fingerprints in this case isn’t the appendage that extends from your hand. Fingerprints are a new form of tracking technology used on the web and you may have heard little about them. Rather than installing a data file tracking data, your device is identified by operating system and a unique tracking key. This means that your use can be tracked across browsers or applications.

So how do we take some control on what data is being collected on us?

There are a few options:

1. Enable private browsing:

Private browsing clears cookies, trackers, and searches at the end of every session.

2. Disable Trackers that integrate into your Operating System and browser

Your OS plays a key role in allowing fingerprint trackers to work. If you disable tracking in your OS this can can help preserve your autonomy from fingerprint trackers.

Here are some places you can do that:

3. Use a VPN

A VPN connection (Virtual Private Network) has two possibilities that you may have heard of. One is to connect to an office’s network remotely, the other way is to connect to VPN’s designed to prevent trackers from finding your location. This also creates inconsistency in their data. A good choice for a free and fast and trusted VPN connection is Opera VPN. One detail about using this method is it can disconnect local network resources while running the VPN connection.

In conclusion, you are a valuable asset to the internet. Although there are many ways to prevent unwanted data from being tracked, tracking is necessary to use any website. It is best to conclude that your value on the internet is best spent on the applications you feel should be benefited by your presence, and making a point that you value your privacy on the internet does effect change on how the internet evolves.

Smart Things versus Privacy

Posted by ABS Tech on November 13, 2017 with 0 Comment

Since the dawn of Internet of things, the tech giants have brought us a wave of new products that fit every nook of our homelife. Among the most revolutionary are smart speakers. Each device is activated by a wake word (“Alexa”, “Hey Siri”, “Ok Google”, “Hey Cortana”) after which it listens for a command or question: “…what’s the weather like tomorrow?” “Order more paper towels”, ” play classical radio on Apple music” “Turn on the lights” are just a few of the types of commands these virtual assistants will answer to. According to Techcrunch.com on a report by Juniper Research, 55% of U.S. households are expected to have a smart speaker. That is 70 million households with a device with microphones capable of capturing sound across a home!

Amazon brought us the Echo lineup, Google brought us Google Home, and Microsoft and Apple recently stepped into the smart speaker world with the Harmon Kardon Invoke smart speaker powered by Microsoft’s Cortana, and Apple’s recently announced HomePod powered by Apple’s Siri. The products function as speakers that can listen for commands, and execute applications and activate other smart home devices.

Amazon Echo 2 is Amazon’s hardware answer for their virtual assistant Alexa. powered by Amazon cloud services (image from Amazon.com)

With this newfound convenience there are concerns about privacy. Google recently revealed that a flaw in the design of the Google Home mini was causing it to record conversations without activation by the wake word. When privacy flaws emerge it does make the average consumer more hesitant to adopt such a powerful device. After all these are devices that were designed to control the rest of the smart things we install in our homes including smart locks, cloud security cameras, and smart thermostats.

Google Home Mini, powered by Google Assistant (Image from store.Google.com)

The concerns many people face today with smart speakers are similar to the concerns we faced at the dawn of cellular phones and in a new wave with smartphones. Suspicions of government or corporate spying on our day to day lives became a concern for some and subject of debate. Since that time, new protocols have been implemented to assure autonomy in personal phone use. Apple encrypts text message conversations in iMessage, and Snapchat launched encrypted picture communications.

Apple HomePod, powered by Apple’s Siri (image from Apple.com)

Based on this history, we can expect a similar evolution to occur as smart things begin to integrate more and more into our daily lives and privacy becomes a weightier concern. already, strides have been made in the programming of these devices to answer privacy concerns Microsoft’s Cortana saves personal data it collects locally to a computer rather than to the cloud in a place called Cortana’s Notebook. If a user chooses to sync data to the cloud, Microsoft provides a highly secure 2 factor authentication on personal accounts. Users can edit or remove data stored there at any time. Apple programmed Siri to hold the personal data it collects encrypted in iCloud storage, so you control what data Apple retains for use of Siri. Google’s privacy site allows you to delete your search history either by voice assistant or by text from Google’s servers at privacy.google.com, all encrypted and secured by your Google account. Amazon has made it clear they wish to prevent exploitation of their users by banning repugnant ads from the Alexa platform .