Jump to content

Welcome to FutureTimeline.forum
Register now to gain access to all of our features. Once registered and logged in, you will be able to create topics, post replies to existing threads, give reputation to your fellow members, get your own private messenger, post status updates, manage your profile and so much more. If you already have an account, login here - otherwise create an account for free today!

These ads will disappear if you register on the forum

Photo

Your 2019 Predictions


  • Please log in to reply
25 replies to this topic

#21
wjfox

wjfox

    Administrator

  • Administrators
  • PipPipPipPipPipPipPipPipPipPip
  • 9,609 posts
  • LocationLondon

 

  • The first terabyte SD card is released

 

*grins*

 

Announced at CES 2019. :)

 

https://www.futureti.../2019/01/10.htm


  • Zaphod, Casey and Yuli Ban like this

#22
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPip
  • 528 posts
I'll start with a vague prediction: Machine Reading Comprehension and Natural Language Understanding will be really big this year.

I can't say exactly what will be produced; but the progress over the past 3 or 4 years has been remarkable -- and lately, it has taken off at an even faster rate. There is so much I could write about this, so much progress I have seen, just in the past year. There is every reason to believe it will spill over into 2019:

* Google's BERT system, for example, has lead to huge improvements in NLU.

* Microsoft Research produced a system called BIGBIRD that is raising eyebrows: https://mobile.twitt...889120386248704

* Are these systems really so hot, are they good at syntax-sensitive dependencies? Apparently, they are: https://mobile.twitt...717916821843969

And several people in this field seem to believe that training even larger models on ever larger amounts of data will lead to further improvements (I believe Jeff Dean hinted at this in one of his talks). So, 2019 might see all these latest systems taken up a notch further, by simply throwing more compute and data at the problem.

I'm not sure what this will mean in terms of products, though, as in-the-lab results don't immediately translate into products. Maybe Google's search engine will suddenly get a lot smarter; or maybe Google Assistant will be even better at answering questions; maybe chatbots will give even better answers; maybe Google Duplex will be even better at handling conversations. All of these have been improving.

I'm unsure what kind of benchmark to give for the prediction.

....

A second prediction is that some major BCI / brain scan group (Openwater and/or Facebook) will release an "alpha kit" product this year. This will not be the BCIs of scifi just yet. They'll probably be expensive and limited in terms of the number of voxels scanned -- but good enough for people to experiment with.

It's possible Openwater will need to delay to 2020 -- I can't say. But it's coming...

By the way, Kernel is now also working on non-invasive BCIs:

https://medium.com/f...ll-1294e30fc93b

"A Vision Shared by Few, Feared by Many, and Destined for All".

....

I could make a prediction about NeuroMod, but I'll refrain.

https://www.futureti...rning/?p=258130

There isn't enough info yet on what they are doing to make a prediction for 2019.
  • Zaphod, Casey and Yuli Ban like this

#23
Guyverman1990

Guyverman1990

    Member

  • Members
  • PipPipPipPipPipPip
  • 756 posts
  • LocationSlocan Park, BC
Japan’s Emperor will abdicate, the first one to do so in over 200 years and this ending the Heisei era.

Following the death of Mahmoud Hashemi Shahroudi, The government of Iran begins to see its influence wane domestically.

#24
Crystal_Gazer

Crystal_Gazer

    New Member

  • Members
  • Pip
  • 1 posts

I'm new here, to the forum, not the website and this site is both a guilty pleasure and place of discomfort - so hello to all the fellow folks like me who spend an inordinate amount of time pondering the future!

 

My predictions for the next decade are:

 

Mass unemployment by 2027 - automation will have made most manufacturing and skilled labor jobs obsolete, forcing the U.S. (and other western nations) to adopt a "basic standard living wage" form of welfare/unemployment.

 

Healthcare industry will continue to boom as humans want to live better and science will continue to see money in this endeavor.  Especially with greedy wealthy babyboomers seeking to live longer.

 

Gen Xers like myself will begin to take over the political and economic landscape and "hopefully" adopt intelligent strategies to combat climate change, wage disparity, crumbling infrastructure, cultural discrimination and incorporating diversity into our society.  White people (which I am) will need to adjust to becoming the minority and make monumental changes to society. White supremacy ideology will begin to fade as millennials become the dominant generation with much more progressive views (THANK GOD.)

 

Climate change, mass extinction, pandemics, scare resources will have to be confronted with cutting edge technology and emerging sciences to create solutions. There is no way to stop it, only guard against it and find scientific solutions.  Desalinization, renewable energy will become prevalent within the next 15 years. 

 

Space mining will become a very important part of our future, as will scavaging massive dump sites for minerals and repurposed materials.  Recycling will become a booming industry as future generations cut down on packaging and post-consumer waste which will be a huge problem for future generations.  I can see in mind's eye companies hired to scavage dumps sites for old electronics as we become more dependent on robots and tech for solutions to our past wastefulness. Dealing with these toxic environmental disasters will be a HUGE business for someone.  Elon Musk? 

 

Food will become scarce in most parts of the industrial world due to climate change, this will cause major conflicts around the world.  Rogue nations will get access to bio-weapons and use as leverage.  The future is exciting, and terrifying.  Science will be the problem and the solution.

 

I encourage everyone to harden to your resources for what's to come in the next 15 years. I'm what they call a liberal prepper.  Check it out.

 

Peace out!


  • Zaphod, Casey and Erowind like this

#25
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPip
  • 528 posts
Related to what I said about Machine Reading and NLU being really big this year, Technology Review just a few hours ago wrote a little blurb about how pretty soon everything will be controlled by voice, thanks to these specific advances:

https://www.technolo...ll-control-you/
 

Whether or not you trust either company with that much control, such a grand undertaking will be limited by what voice assistants can understand. And compared with other subfields of AI, progress in natural-language processing and generation has kind of lagged behind.

But that could be about to change. Last year several research teams used new machine-learning techniques to make impressive breakthroughs in language comprehension. In June, for example, research nonprofit OpenAI developed an unsupervised learning technique to train systems on unstructured, rather than cleaned and labeled, text. It dramatically lowered the costs of acquiring more training data, thereby increasing their system’s performance. A few months later, Google released an even better unsupervised algorithm that is as good as humans at completing sentences with multiple-choice answers.

All these advancements are getting us closer to a day when machines that really understand what we mean could render physical and visual interfaces obsolete—and usher in the full potential of an OMO world. For better or worse.


I didn't mention OpenAI -- I should have, and should have also mentioned AllenAI, which has contributed to these recent advances. And it's all very recent. In previous years, there were big advancements, with systems reaching human-level performance on specific reading benchmarks; but nothing like what has recently come out of Google, Microsoft, OpenAI, AllenAI, etc.

Still, AI trained with brain data would take that up a few more notches, enabling system that not only do a good job of getting the gist of things; but understand it like a human, including a lot of the commonsense and pragmatic subtlety of language. They will not just be able to fake "human" (like Google Duplex, and the even better system by Alibaba) by hiding what they don't understand, and steering the conversation; but will actually understand.
  • Yuli Ban likes this

#26
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPip
  • 528 posts

Synthesized 3D objects you can view from different angles in VR (3D VR still images) may be a big thing in 2019. There have been some impressive advances lately:

 

https://venturebeat....nd-reflections/
 
https://papers.nips....representations
 
Note that this is much harder than synthesizing 2D images. And 3D VR video is even harder, still.
 
Also, GANs have recently been shown to learn things about the world that border on "commonsense":
 
https://www.technolo...ust-like-we-do/
 

It’s been unclear before now whether there was any way of learning this kind of thing... That it *is* possible suggests that deep learning can get us closer to how our brains work than we previously thought...


So, the 3D image synthesis might include details you might not expect machines to be capable of producing.

I think various groups are looking to "the next big thing" in GAN-based media synthesis, having shown they can get good results for 2D stills. 2D video synthesis is one possible direction; but 3D VR scenes will probably be easier, so they way focus their effort on that in 2019. I expect to see several more results like the above, perhaps for whole scenes, instead of lone 3D objects.


  • Yuli Ban likes this




1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users