Why Robots Mean Interest Rates Could Go Even Lower In The Future

[Reginald Noble]

What is it?

I have to interpret and analyze web and 3rd party data then optimize marketing campaigns based on it.

It'll take another 100 years to even come close for a machine to be able to do what I do - And even then a human will have to QA the data sets as well as the machine's logic on the regular basis. Automation can never be perfect, there are a million and one things that can go wrong in any technology based process or project.

 

[Yogi]

I have to interpret and analyze web and 3rd party data then optimize marketing campaigns based on it.

It'll take another 100 years to even come close for a machine to be able to do what I do - And even then a human will have to QA the data sets as well as the machine's logic on the regular basis. Automation can never be perfect, there are a million and one things that can go wrong in any technology based process or project.

That sounds like it could be automated.

 

[Reginald Noble]

That sounds like it could be automated.

It would sound like it to someone unfamiliar with the industry and digital marketing metrics in general. They already tried in the early 00's (right after the inception) and it failed miserably.

They couldn't even outsource the job because Habib Al-Haram couldn't pronounce "Click through rate". Let alone glean any actionable insights.

 

[Reginald Noble]

I also forgot to mention as long as API connectivity remains sub-optimal and black boxed between applications, humans will especially be needed. We have problems with data parsers ALL THE TIME because they break due to exceptions and can't ignore or navigate around ambiguous data. This is where the human comes in to provide guidance.

Most data sets are a mess. Machines can only be programmed to generally operate off predictable circumstances. "Predictive" modeling is based on core inputs from humans. The machine itself cannot initiate predictive modeling at all.

And even if we migrate to more of an open source format humans will still be needed to facilitate the "translation" of the data from one application to the next. Machines cannot do this because they don't understand what we find important in general from a business perspective.

Business needs change every 90 days based on capitalism and shareholder maximization principle. Machines simply can't keep up or process all the "data" that goes into what a shareholder finds important from one quarter to the next.

Mark Cuban would have said something if he thought IT and other tech could be fully automated anytime soon.

 

[BaggerofTea]

I also forgot to mention as long as API connectivity remains sub-optimal and black boxed between applications, humans will especially be needed. We have problems with data parsers ALL THE TIME because they break due to exceptions and can't ignore or navigate around ambiguous data. This is where the human comes in to provide guidance.

Most data sets are a mess. Machines can only be programmed to generally operate off predictable circumstances. "Predictive" modeling is based on core inputs from humans. The machine itself cannot initiate predictive modeling at all.

And even if we migrate to more of an open source format humans will still be needed to facilitate the "translation" of the data from one application to the next. Machines cannot do this because they don't understand what we find important in general from a business perspective.

Business needs change every 90 days based on capitalism and shareholder maximization principle. Machines simply can't keep up or process all the "data" that goes into what a shareholder finds important from one quarter to the next.

Mark Cuban would have said something if he thought IT and other tech could be fully automated anytime soon.

True if you are just looking at automation being a machine following through a set of pre-programmed instructions

However artificial intelligence and the ability to develop machine intellect will throw all of that out of the window

 

[Reginald Noble]

True if you are just looking at automation being a machine following through a set of pre-programmed instructions

However artificial intelligence and the ability to develop machine intellect will throw all of that out of the window

nikka I'm gonna have to go deep on this:

Unfortunately we are still thousands of years (if that early) away from even basic AI being usable in the real world.

If you look at the Japanese androids made they STILL have predictable emotional expressions and responses to everyday stimuli. The reality is humans everyday have unpredictable reactions to everyday stimuli AND our reactions evolve as continuously exposed.

I just can't see AI being able to replicate our intelligence which is factored in by things you cannot simulate such as hormone level fluctuation and most importantly DNA evolution.

The key to developing an android which is most human like is to simulate the mitochondria itself from a cybernetic perspective. I imagine that would require increase in nanotechnology which is extremely expensive.

It will be light years before humans can enable an artificially created being to have something as an everyday simple thing we take for granted: an orgasm...

It requires hundreds of hormones produced, and all in synchronization. You have to replicate the internal organs of a human, the brain alone won't be enough.

Another thing is a lot of business decisions are made on primal instinct rooting back millions of years. This data will be nearly impossible to port over into any AI being and they are essentially starting from scratch, the beginning of "their" evolution. How can we overcome that gap?

 

[BaggerofTea]

nikka I'm gonna have to go deep on this:

Unfortunately we are still thousands of years (if that early) away from even basic AI being usable in the real world.

If you look at the Japanese androids made they STILL have predictable emotional expressions and responses to everyday stimuli. The reality is humans everyday have unpredictable reactions to everyday stimuli AND our reactions evolve as continuously exposed.

I just can't see AI being able to replicate our intelligence which is factored in by things you cannot simulate such as hormone level fluctuation and most importantly DNA evolution.

The key to developing an android which is most human like is to simulate the mitochondria itself from a cybernetic perspective. I imagine that would require increase in nanotechnology which is extremely expensive.

It will be light years before humans can enable an artificially created being to have something as an everyday simple thing we take for granted: an orgasm...

It requires hundreds of hormones produced, and all in synchronization. You have to replicate the internal organs of a human, the brain alone won't be enough.

Another thing is a lot of business decisions are made on primal instinct rooting back millions of years. This data will be nearly impossible to port over into any AI being and they are essentially starting from scratch, the beginning of "their" evolution. How can we overcome that gap?

You are making it more than it needs to be. All you need is artificial intelligence that can adjust itself to randomized variables. "Emotional judgment" can be reserved for a human overseer.