cancel
Showing results for 
Search instead for 
Did you mean: 

Within 15 years, VR will seem 100% real, says THIS guy...

cybernettr
Superstar
One of the appeals of this guy, whose book I've read and who I've been seriously following for 5 or 6 years now, is his prediction that VR would be a "thing."

He stood by this prediction in the face of the general belief until fairly recently that VR would forever remain a fad of the '90s. I know all this, because this was the general consensus on VR only about 5 years ago, when I first started following the issue.

Now this guy says that within 15 years, VR will seem 100% real and will incorporate human-level AI for the creation of real-time endless virtual worlds with characters you can interact with that will seem almost as real as real reality.

Short video: http://www.cbc.ca/player/play/2682488530
77 REPLIES 77

cybereality
Grand Champion
That video was from a year ago, so looks like we only have 14 years to wait...
AMD Ryzen 7 1800X | MSI X370 Titanium | G.Skill 16GB DDR4 3200 | EVGA SuperNOVA 1000 | Corsair Hydro H110i Gigabyte RX Vega 64 x2 | Samsung 960 Evo M.2 500GB | Seagate FireCuda SSHD 2TB | Phanteks ENTHOO EVOLV

cybernettr
Superstar
Huh, the link I followed said it was from January of THIS year, not LAST year. In any event, sorry for posting old news...only 14 years in advance! LOL  😄

On the plus side, it's not like we have to wait 14 years to reap the benefits...VR is going to get a little better every year. Something to keep in mind when you hear those who STILL say it's just a fad (and there are still plenty of them out there). Also, it's nice to know what to look forward to.

cybereality
Grand Champion
I love Ray Kurzweil and have read his most popular books. He's probably right, even though a lot of people doubt him I think we will all be surprised by the future. I don't mind that the video is old, I hadn't seen it before and it's still relevant. Thanks for sharing.
AMD Ryzen 7 1800X | MSI X370 Titanium | G.Skill 16GB DDR4 3200 | EVGA SuperNOVA 1000 | Corsair Hydro H110i Gigabyte RX Vega 64 x2 | Samsung 960 Evo M.2 500GB | Seagate FireCuda SSHD 2TB | Phanteks ENTHOO EVOLV

Sax-a-boom
Rising Star
I'd like to see his reasoning and numbers on this because some of those claims are way out there. Render farms can't truly convince us that something created in CGI is indistinguishable from reality. Not going to happen in realtime on your computer in 15 years. Talking of which his claim about processing power being a billion times greater in 25 years is a bit misleading because of the caveat he makes of 'per dollar'. Obviously costs tumble per calculation as time goes on, but he gives no indication of how much more processing power we'll have in 25 years. Improvements have noticeably slowed over recent years and unless there's some radical breakthrough, years between cycles are going to lengthen each generation.
Also what does computers will be 100,000 times smaller mean? What is he defining a computer as? Again if he's talking about the processor, that seems like an impossible claim. Processor shrinks mean we are now down to 10nm, only 20 times wider than the width of some atoms. So 100,000 times smaller is going to need some new physics.

Anonymous
Not applicable
While I have no expertise in the field... human level AI in 15 years? I don't believe that. It's not like AI has come all that far in the last 15 years. They still haven't got decent natural language recognition, let alone natural language understanding, and while visual recognition and use of vast data sets has come on by leaps and bounds, they haven't progressed so far in the last 15 years that it seems likely that they'll reach the levels of human capability in the next.

And I haven't heard that cognition and conversation have come all that far. Don't most chatbots rely on conforming to stereotypes, and exploiting the extremely superficial and ignorant posting habits of some people's online communication?

Happy to learn otherwise, of course, but I think that that this would be one of Mr. Kurzweil's predictions that will end up being filed under 'inaccurate'.

agenttoff
Heroic Explorer
People always make ridiculous predictions like this, it's one of my pet peeves.  "Neural implant VR in 10 years!!"  15 years is such a blip.  

cybernettr
Superstar

AndyW1384 said:


Happy to learn otherwise, of course, but I think that that this would be one of Mr. Kurzweil's predictions that will end up being filed under 'inaccurate'.


 Even though I don't think much of Business Insider, that was actually a pretty good analysis.  But even with those predictions Kurzweil is "wrong" on, he's with in the right ballpark (I'm dictating this reply right now on my iPad--I do most of my extensive text entry through dictation).  So even if Kurzweil is over optimistic about something by about 10 years, if he's predicting something that totally revolutionizes the whole world, I would have to give them credit for being overall right. 

cybernettr
Superstar

AndyW1384 said:

While I have no expertise in the field... human level AI in 15 years? I don't believe that. It's not like AI has come all that far in the last 15 years. They still haven't got decent natural language recognition, let alone natural language understanding, and while visual recognition and use of vast data sets has come on by leaps and bounds, they haven't progressed so far in the last 15 years that it seems likely that they'll reach the levels of human capability in the next.


 Don't look to chat bots for the direction of natural language processing, look to something like IBM Watson.  This showed very good natural language understanding, since it won Jeopardy by reading Wikipedia and scores of other natural language documents.  

Kurzweil has been very consistent in saying that human level AI will be here by 2029.  In fact, he has a long-standing bet with Mitch Kapor, founder of Lotus.  The test he's predicting won't be a wimpy test, either – it will be something like a two hour long marathon with testers who know what they're doing, not some dumb celebrities or people taken off the street. 

cybernettr
Superstar


I'd like to see his reasoning and numbers on this because some of those claims are way out there. Render farms can't truly convince us that something created in CGI is indistinguishable from reality. Not going to happen in realtime on your computer in 15 years. Talking of which his claim about processing power being a billion times greater in 25 years is a bit misleading because of the caveat he makes of 'per dollar'. Obviously costs tumble per calculation as time goes on, but he gives no indication of how much more processing power we'll have in 25 years. Improvements have noticeably slowed over recent years and unless there's some radical breakthrough, years between cycles are going to lengthen each generation.


I don't know about that. 15 years ago, we had this, and now we have project cars and dirt rally: 

yjvk3rdbbguo.jpg