Here's what life will be like in 2022 according to IBM

[Ripster42]

Flying cars that you get to drive are never coming. The only way that would happen is if it was completely autonomous and there was no way you could gain control of the vehicle. Sorry guys, terrorism exists.

[twiddler]

Flying cars that you get to drive are never coming. The only way that would happen is if it was completely autonomous and there was no way you could gain control of the vehicle. Sorry guys, terrorism exists.

Man, terrorism is the least of our worries. Humans...

[Chickat]

In the 80s, scientists predicted we will have flying cars in 2015.

It didnt happen. I don't believe development will be that fast.

I believe it could have happened though. We just didn't want it enough. I also believe we could have had a perm base on Mars by now if we really wanted it as a species.

Anyways, most on the list is probably 10-15 years away, not 5.

[Celista]

1. unlikely, there will be a lot of ai in 5 years but speech analysis on that level is a stretch, if only due to the speed at which data the ai can learn from can be gathered. on top of that it would already have to be in FDA approval procedures to make the cut.
2. unlikely, too advanced/expensive for 5 years unless the military had it right now, and if the military had it right now they wouldn't be allowed to talk about it.
3. sure, military has cameras that can do that right now and has publicly talked about it, so possible.
4. already exists, so possible it will reach mass market in 5 years. basically just a glorified name for a smaller pregnancy test/blood sugar test/urine analyzer that can do more complicated tests.
5. sure, basically just evolution of existing tech, unlikely to reach high saturation/deployment levels in 5 years though unless they start real soon.

Basically what I was going to type. #1 and #2 are real stretches, I don't see #2 coming about to mass market for another 20 years to be honest. That's accounting for the speed-up of technology.

[Zandalarian Paladin]

IBM tend to be really creative in their predictions. Quite frankly, it is mostly done to hype the technology and get public attention.

Now that doesn't mean it's bound to fail. The past five years, we've seen Big Data and learned how to exploit it. Now we're at a point where Big Data can't go much farther -- abstraction, or in other words the human input, prevent Big Data to accurately make "intelligent" steps. The ability to improvise and adapt on an individual level.

Here's where things gets interesting: Deep Learning. The new toy that is bound to revolutionize how our technology, market, politics, science, medicine and entertainment. As Wikipedia puts it:

Deep learning (also known as deep structured learning, hierarchical learning or deep machine learning) is a branch of machine learning based on a set of algorithms that attempt to model high level abstractions in data. In a simple case, there might be two sets of neurons: ones that receive an input signal and ones that send an output signal. When the input layer receives an input it passes on a modified version of the input to the next layer. In a deep network, there are many layers between the input and output (and the layers are not made of neurons but it can help to think of it that way), allowing the algorithm to use multiple processing layers, composed of multiple linear and non-linear transformations.[1][2][3][4][5][6][7][8][9]

More specifically: allows the computer to make decisions through a complex neural network. To think.

[Kaleredar]

In the 80s, scientists predicted we will have flying cars in 2015.

It didnt happen. I don't believe development will be that fast.

I don't know if Back to the Future 2 was meant to be taken all that seriously.


Though they did essentially predict Trump...

[shise]

This reminds me of back in the early 90s there were predictions that by early 2000s we would have flying cars.

And so we did. But they proved to be useless.

- - - Updated - - -

In the 80s, scientists predicted we will have flying cars in 2015.

It didnt happen. I don't believe development will be that fast.

We have flying cars for nearly 10 years, they just happen to be useless. So instead we have self driving Tesla.

[PC2]

Deep learning in its current state isn't a black box you're making it out to be. You still have to be very conscientious of your inputs, parameters, etc. and it still requires that "human input" for validation, feature engineering, etc.

The black box is referring to the inability to understand the representations. Most all of the big neural nets are like black boxes because the data representation goes toward being decorrelated. Sometimes they aren't blackboxes though. Like with visual facial models you can see how they create edges and facial features on some layers. AlphaGo on the other hand would give you little clue as to how it is arriving at a solution, outside of the optimization result it spits out.

[May90]

I always expected flying cars to come after automated ones. People suck enough at 2D driving, I could only imagine what 10,000 cars in rush hour traffic would be like under human control with the vertical axis in play.

Pretty sure at that point the control will be handled by complex AIs, and the humans will only have to specify the destination.

[TheWorkingTitle]

..

ever hear this version?

[Zandalarian Paladin]

Deep learning in its current state isn't a black box you're making it out to be. You still have to be very conscientious of your inputs, parameters, etc. and it still requires that "human input" for validation, feature engineering, etc. And that ignores the computational problems such as scaling.

Indeed. But I haven't made the claim that in its state, it is perfect. The good thing about DL is that once we've been able to develop a solid frame, we should be able to almost exponentially iterate and progress. But we yet have to make a convincing frame.

Much like Big Data, in fact. Given a few years, I have no doubt Deep Learning will be the next big thing.

[ItachiZaku]

Meh, seems kinda weak to me.

[Halicia]

I want my hover board first.

along with flying cars, hover boards already exist.

[Kuntantee]

IBM tend to be really creative in their predictions. Quite frankly, it is mostly done to hype the technology and get public attention.

Now that doesn't mean it's bound to fail. The past five years, we've seen Big Data and learned how to exploit it. Now we're at a point where Big Data can't go much farther -- abstraction, or in other words the human input, prevent Big Data to accurately make "intelligent" steps. The ability to improvise and adapt on an individual level.

Here's where things gets interesting: Deep Learning. The new toy that is bound to revolutionize how our technology, market, politics, science, medicine and entertainment. As Wikipedia puts it:



More specifically: allows the computer to make decisions through a complex neural network. To think.

Neural Networks are old, probably more than 50-year-old. We've been exploiting it for a long time, especially after computational power of machines got to the point which met the requirements of neural systems. Deep Learning is a step forward, but hardly revolutionary.

Your perception and maze are solely constructed by the way Deep Learning is advertised in media, i.e., popular science.

For early neural systems, you can read about Hodgkin-Huxley model.

[PC2]

Neural Networks are old, probably more than 50-year-old. We've been exploiting it for a long time, especially after computational power of machines got to the point which met the requirements of neural systems. Deep Learning is a step forward, but hardly revolutionary.

Your perception and maze are solely constructed by the way Deep Learning is advertised in media, i.e., popular science.

For early neural systems, you can read about Hodgkin-Huxley model.

Yeah they are old, but as long as you keep getting more computation and good data they just keep getting better because gradient descent scales. There is still a lot left to get out if them through better GPUs, then probably FPGAs and ASICs after that.

[Deleted]

Five years is way too soon for such a tech leap.

[Drekker17]

None of this will happen haha, science predictions are always wrong, always.