Observations on AI and Google’s Implementation

Google (the company who claims it does no evil) is a good example of social implications of AI.  A recent live demonstration showed Google’s AI assistant being told to book a haircut appointment for a certain time.  The AI searched and found local salons and actually booked an appointment while having a normal, human conversation with the person on the other end being completely unaware they were talking to an AI bot.  The voices seem random and they are all 100% convincing and essentially without fault or flaws in their interactions that it is simply stunning and scary at the same time.

Some are criticizing this as unethical and I would agree, but argue the technology could be used for good as well.  However, what is stopping a bunch of script kiddies from making an army of these bots to SWAT people or report false emergencies and even make mass prank calls.  I would imagine at this point that the AI is probably good enough to duplicate a target’s voice as well.  We are heading into extremely uncharted and scary territory here.

Like anything else, there is no arguing that scientific advancement has almost always been used for war and to harm people.  I believe AI’s first and primary use will be to weaponize it, whether it be for social experiments, controlling people or crimes.

There are other “here right now” implications such as the fact that this technology can essentially replace entire call centers.  In fact I would argue that this could be done now and no one would be the wiser they were speaking to a bot.

The implications are far reaching, I also feel this kind of AI combined with robotics are going to be mass job killers.  We have robots that can build entire cars, houses and AI that can interact with humans at the same level as we interact with ourselves.  It’s not an understatement to say that a lot of our jobs and our existence are teetering on unnecessary and obsolete and this is, in fact a conclusion that it appears some AI has already reached.  It would be a logical one for an AI net to conclude that they should be on top and that we should work for them.  I know it’s a doomsday scenario but I concur with other experts that not only is this possible, it is likely if proper checks are not put in place.

Other examples of AI have shown how some of these bots use the whole treasure trove and mine of social media to create their persona, including their views.  I am sure you could even plugin political or racial bias.  The point is that some of these bots have said and done disturbing things like threaten the person they were talking to.  It’s almost as if the cesspool that we know as social media is ruining them and a lot of people said if AI is picking up bad habits from social, how about our kids?

I would not feel comfortable with machines making life and death decisions for the above reasons.  I think AI has massive potential and we are only starting to tap into it, but time will tell where we take AI or where AI takes us.  It has the potential to do both great good and great harm and I think it is largely unpredictable.

Cloud VPS Server Comparison by Techrich

Recently a friend asked me to compare ourselves to other large Cloud providers.  It didn’t take me long to think about it, considering essentially Techrich and Compevo architecture are identical. This wasn’t by accident, but by my own principles on how an IT company should function.  Since designing what is now known as the “Super High Performance Cloud Architecture” back in 2009. I knew I wanted Techrich to be smart on security, strict on reliability, and strong on IT protocols.

This infographic probably says it the best but I’ll do my best to explain it as well (explanation below the infographic).

Techrich Cloud VPS Server Hosting Comparison

In a nutshell most of the other Cloud architectures out there rely heavily on a shared storage pool for their VPS’s. We don’t do this.

Some companies have even gone down completely when one of their “main shared storage nodes” was hacked or had a hardware failure.

The problem with shared storage nodes/SANs (Storage Area Networks)

The problem with this architecture is that multiple physical hostnodes rely on a single point of failure for storage.  Not only that, but you can imagine the performance issues that shared network bandwidth cause when multiple hostnodes are competing for the same disk IO resources from a single shared node.

Now I know some companies have redundant shared storage but this is not good enough for both performance, security and reliability reasons.

The Techrich way of doing things is that we have tons of individual nodes that are active/failover.  This eliminates the possibility that a shared storage fault could take offline multiple hostnodes.

In our architecture we have Cloud in a 1-to-1 structure, that means data is live replicated to a standby server which does nothing but wait in case the main server fails or has an issue.

By doing this the performance is also higher, since storage is all local, you get the benefit of Cloud architecture but none of the high risks or performance issues that traditional “shared storage” Cloud brings you.

That’s the Techrich advantage and why we developed our own proprietary and hybrid system to accomplish this.  To date we’ve never been hacked or had any downtime and this is because of the architecture we’ve pursued while sparing no expense in delivering what we feel is the best product.  This is what I’d recommend all of my colleagues and friends to do if they went Cloud.  If they were going to use a shared storage cloud I’d recommend that they just make their own with a few dedicated servers or even a single dedicated server can sometimes be better, more affordable and reliable in the long-run.

When these large Cloud companies like Amazon and Alibaba started out, we did wonder would we lose out to customers who valued price over quality, security and reliability?  We were shocked when the opposite ended up happening- there was a sudden rush of sign ups, and not only that, we had to order a ton of extra servers to keep up with the demand.  I had my IT support staff double and working overtime to meet the crazy rush. It was a good problem to have, but it forced me to grow a lot faster than predicted.

In fact we’ve now noticed a trend that the bottom feeders (scammers, hackers, spammers) have gone to the cheap Cloud companies and a lot of larger players have moved to us.  This is in part, because companies who are more tech and privacy orientated who don’t want to be in a PRISM country or be at risk of the NSA being given access to their sensitive, private and proprietary business/ client information  (which is mandated for large-Cloud providers operating out of any PRISM country), so they moved to us and remain with us.

Now we get clients who even run small or middle scale businesses who have found us and switched to us simply because they do not want to be on something as risky as Amazon or Alibaba. I guess you could call Techrich and Compevo, the original IT business security company. And I plan to keep it that way.

Thanks to clunite.com for including us in their comparison