cancel
Showing results for 
Search instead for 
Did you mean: 

The Oculus Summer Sale spawns a new competitive analysis

Zenbane
MVP
MVP
A great new article has been published by VRBeginnersGuide.com

Throughout the past 15 months there has been ongoing debate centered around the competition in the VR Industry. At the forefront of that debate is, of course, the quintessential: Rift vs Vive

Everyone has their opinion, bias, and speculation. The lack of "official sales" numbers also forces consumers to rely on alternative sources of information (e.g. surveys, research agencies, search engine statistics).

Yet the Oculus Summer Sale has dropped the Rift down to a price that was promised before its official release (In September of 2014, Palmer went on record for stating that the Rift should reside between $200-$400). Thanks to the lengthy Summer Sale, the competitive analysis has returned to the public eye.

I love the VRBeginnersGuide.com article because it addresses the full spectrum of the Rift/Vive debate, and paints a very clear picture of a reasonably expected outcome.


Is this the Beginning of the End for the HTC Vive?





Things were looking rather grim for Oculus at the end of 2016. SuperData Research estimated there had only been 243,000 Rift units sold verses 420,000 HTC Vive units. Even with Oculus releasing their Touch controllers near the end of the year, it seemed like much of the community had already written Oculus off. The new controllers were supposed to bring the Rift into direct competition with Vive’s Room-scale tech, but launch did not go smoothly. Full room-scale was still officially experimental and thousands of users reported glitches with the tracking.

At first glance, it may seem the Vive is going from strength to strength. Releasing some cool new trackers to bring more objects into the virtual world, and the Deluxe Head Strap. There is an eye tracking attachment in development that will allow foveated rendering and the new Knuckle controllers are set to be the new standard with full 5 finger tracking.
But, all of these new upgrades have issues.

The trackers will, at best, result in a whole bunch of new peripherals that are only suited to one or two jobs. Eye tracking is great for developers working on foveated rendering, but from a consumer stand point it probably won’t see widespread use in games until a year to two’s time. By then we will be staring down the barrel of VR generation 2. The Deluxe Head Strap will set you back at least $100USD and early adopters found the foam on the back disintegrated in contact with water. Reviews indicate that at best the comfort level only matches that of the Rift. So, that’s $100 to bring an already more expensive product up to the same comfort level as the competitor?

Finally, while the Knuckle controllers looks awesome, they are made by Valve, not HTC. Let’s talk about why this is a bad thing.

Not everyone realizes that the Vive was actually a joint venture between Valve and HTC. Valve developed Open VR and the Lighthouse tracking system, and Vive builds the hardware. So, for a start the main trump card HTC has, Lighthouse tracking, is not even their tech. The upcoming LG headset, which will shortly be followed by other major brands, will use the exact same tracking technology. The Knuckle controllers will also most likely be compatible with these new systems.

The main implication of this arrangement is that HTC only makes money from selling hardware, the Vive and it’s add-ons. Steam wanted VR to be an open platform and prevented HTC from developing exclusive software to go with the Vive. Recently HTC circumvented this agreement by opening Viveport, an exclusive software environment only accessible on the Vive, similar to OculusHome. But Viveport can only save them if they continue to dominate the market. That will be difficult for two reasons. 1: Other major brands are about to jump aboard the Steam VR train, and  2: Oculus price cuts.


Why You Won’t See a Vive Price Drop

Oculus has a different business model, the Software ecosystem. Their model is all about making money from selling software for the Rift, not selling the Rift itself. This is why Oculus was able to drop the price so drastically less than 12 months after the original release. The more Rifts in homes, the better. HTC on the other hand famously and proudly proclaimed they were selling the Vive for a profit from launch. This would explain not only why many Vive users experience broken hardware (dead pixels, broken lighthouses, shoddy controllers and disintegrating head straps), but also why HTC will never be able to price match with Oculus.


HTC has Been in Trouble for Years



This is what has happened to the HTC stock price in the past 5 years. It’s not a pretty picture. In August of 2015 their stock price was so low that the company was valued at less than the value of their actual physical Assets. A signal that investors have zero faith in the company’s long-term earnings potential, either in the form of eventually turning a profit or even getting acquired by a larger firm. The price has barely shifted since then. They recently had to sell one of their factories just to keep up funding for VR and with their latest smart phone failing to score a decent slice of the market there is little reason to believe things will get any better any time soon.


Signs of the End

We quoted SuperData’s sales estimates at the start of this article. 243,000 Rift units verses 420,000 Vive units. The more recent numbers, however, are looking quite a bit different. A June 2017 report by Research firm IDC estimates the Rift has sold about 520,000 units, compared with 770,000 of HTC’s Vive. Which this does show the Vive still in the lead. Have a think about what these new numbers mean. If both estimates are accurate, then in the past 6 months Oculus has sold 277,000 to HTC’s 350,000. Much closer numbers than last year. Also, Oculus just announced the new all-in-one Rift package will be set at a price of $499USD, close to half the price of the Vive. With both systems now performing practically the same; how do you think those numbers will look in another 12 months?

We have little doubt that VR is here to stay. The overall numbers just keep getting better and VR’s future is looking to be on very solid ground this year. But with LG, Microsoft, Apple, Acer and ASUS among others about to enter the market, the Vive’s days as Steam VR’s poster boy look well and truly numbered.


Full article:
https://www.vrbeginnersguide.com/is-this-the-beginning-of-the-end-for-the-htc-vive/

One source reporting news of HTC selling their second Factory:
https://www.androidheadlines.com/2017/03/htc-selling-another-phone-factory-fund-vr-business.html

❤️
43 REPLIES 43

Zenbane
MVP
MVP


AR and VR are on a downturn as most investments are now shifting towards AI development.


I actually consider that extremely valuable progress. Artificial Intelligence and Machine Learning have been around for a very long time yet have not been able to prove themselves as a true asset in the workplace. In order to make them valuable today, human intervention and manual oversight is needed, which defeats the purpose of artificial intelligence and machine learning in the first place.

We need to remove the manual human-factor from A.I. and then integrate the innovation with AR/VR.

Today the best we can do in the corporate environment with AR/VR is expand upon things that exist today: charts, graphs, datagrids, and virtual classrooms/meetings.

When we can get A.I. to function on its own and mix it with AR/VR, then we will have hit yet another societal evolution; as opposed to the current state where we're simply re-inventing existing standards and practices.

A.I. first, VR second... and THEN we level up to the 3rd Matrix Film:




'Cause right now we're stuck somewhere between 1 & 2, na' mean?

Anonymous
Not applicable
mmm I wouldn't say we need to make AI a top priority just yet. Sure - it be cool for it learn and have games grow with us - but at the same time - we're no where near the point that AI in games really matter just yet. We are still years out from the VR dream of a world around us.

Full body hapics still need to become a thing, taste, smell, body tracking, some way to move with in a limit space, eye tracking, increase visual with limited hardware, etc etc

Some of this can happen sooner than later for sure and some of this stuff we're really close cracking either way, but there still a big gab between when we hit a need for AI that smart to what human manual can do for us right now as well still.

Some basic AI and full body tracking could already create some massive skill level increase alone for fighting games or RPG games. Long as the AI knows what you are about to do just like any other human would watch your body langue - the harder it will be for a human to out skill even basic AI control.

Right now, all AI is base on action reaction - the problem with that is it has to wait to know what to do next and that makes it slow - What we need in stead of action reaction do sight - planning - action - sight - reasoning - reaction. Sighting being what are we doing and what is going on around the world with it. Planning is all ray track that can happen in real time basic off stats - and then finally the action/animation. If you are about to slash the AI with a sword:
AI sees that you lifted your hand - AI plans by weighing the actions of stepping backwards in time or defending if it can or random change of "fear" (roll a dice) - does action it think is best - takes another look around for anything that might be helpful or what action it can do next - plan next action stab or run away  or use items around - reaction on player.

As for AI on other things - it makes sense that AI might be getting some traction. Hardware is finally playing cache up on the ideas and works of other software and that means they are getting more powerful than what a human can do with in a given amount of time. That AI isn't the same as real AI. It will never take over the world even if we like to think of skynet:) It's more ... chart base pattern seekers.

With that said - I don't think AI will ever be better than humans. At the end of the day we are all just rolling dice and hoping we don't get 1s each move we take. AI won't be any better than we are and skynet would know that. If human have problems driving cars and we been doing it for a while now - how do we expect AI base driving to really be 100% safe? Sure it can be done - but that means reworking roads and everything else. AI and humans both know that is going to cost a ton of resources and time to rework the road system to be better managed. This means both AI and humans have to take risk giving in a way that makes sense. Look Before You Leap in this case by checking senor data first - then moving forward all while hoping 100% of the time nothing bad will happen base of the collected senor data. If not enough senor data is collected - sure there will be problems, but that is a risk everything has to take when we move forward or backwards in any situation.

Zenbane
MVP
MVP
In the gaming Industry, A.I. is making big strides with breakthroughs in games that not only involve memorizing multiple combinations simultaneously (e.g. Chess), but also with games that require intuition:

Google's AI Wins Fifth And Final Game Against Go Genius Lee Sedol
In the final game of their historic match, Google's artificially intelligent Go-playing computer system has defeated Korean grandmaster Lee Sedol, finishing the best-of-five series with four wins and one loss.

The win puts an exclamation point on a significant moment for artificial intelligence. Over the last twenty-five years, machines have beaten the best humans at checkers, chess, Othello, even Jeopardy! But this is the first time a machine has topped the very best at Go—a 2,500-year-old game that's exponentially more complex than chess and requires, at least among humans, an added degree of intuition.

https://www.wired.com/2016/03/googles-ai-wins-fifth-final-game-go-genius-lee-sedol/


However, the focus of what I referred to is less about gaming/entertainment and more about industry and social media. If a Corporation can have a piece of A.I. software monitoring all activity (sales, profits, loss, revenue, asset valuation and devaluation, depreciation, expenses, cost of operations, employee performances and salary, competitive analysis, market fluctuations, etc), and then use all that to make accurate predictions about an organization's expected performance and even Stock behavior... then holy crap we have evolved in a way that will skyrocket Capitalism and Enterprise to heights never dreamed of before.

Right now all this stuff requires millions of dollars in payouts to I.T. staff: Data Architects, Data Scientists, Business Intelligent Developers, and Software Programmers; plus licensing for the software, and operational costs for the hardware.

And all that money just to create something that might be accurate 50-80% of the time, at best. Worse yet, predictable analysis (which is the dream we want to achieve with A.I.) is far from real-time today. By the time humans aggregate everything and present it in a meaningful way, a multitude of internal and external events have taken place that could change the outcome.

When a piece of A.I. can analyze everything up to this exact moment in time and Output a variety of decisions with precise outcomes:




... and do it all in a matter of seconds with up-to-the-minute data, then AR/VR will simply put a face on top of everythig; a face that we can talk to, or information we can hold in our hands.

Advances in A.I. are the key to making HMD's a standard for daily headgear  🙂

Anonymous
Not applicable
Mmm I guess I understand a bit too much on how some of these shortcuts can work with in AI learning. Games are a really bad example of AI learning and mastering of anything. Because - well - games all have outcomes that can be ray trace down to a single point of action. So long as the AI knows the rules - they just follow a flow chart tree focusing down the path that leaders to a victory in most cases. Much like the image you was showing.

Even talking AI can learn speech already and new words already. Why we haven't seen these in games yet - well I have a feeling it comes more down to the dev not wanting to create an AI for the world mostly:) Problem with AI is just that it can pick up the bad points of us playing with it as well. Like telling it - it's ok to drink battery acid when it's totally not to xD

"sales, profits, loss, revenue, asset valuation and devaluation, depreciation, expenses, cost of operations, employee performances and salary, competitive analysis, market fluctuations, etc), and then use all that to make accurate predictions about an organization's expected performance and even Stock behavior" -- Yes this is where AI will shine for sure. This is the real test for AI and how well it can predict outcomes base off so much senor data coming in. Here - there isn't a lot of short cuts you can take to make this work well and this is where something learning how everything flows, delays, and works is a bigger undertaking than playing a game for sure!

Right now, I don't see why game AI isn't at a higher level anyways for game/software as it is. Maybe it's not something we're lacking from a understanding point of view - but an asst much like all the objects we see in a 3D/VR world. To me - it just feels like we're lacking the open source code that game engines need to work on more than anything at this point. I don't feel like it is a priority from a hardware stands it is needed - but from a open stander point of view we need to look into it more.

Maybe it is limited by hardware - but I see soo many projects still using older hardware making some big progress no game really takes advance from.

Morgrum
Expert Trustee
Zenbane can we at least wait until im old and grey before we create skynet!
WAAAGH!

Anonymous
Not applicable
Nice night for a walk!
Wash day tomorrow. Nothing clean, right?

 😄 

Anonymous
Not applicable

Morgrum said:

Zenbane can we at least wait until im old and grey before we create skynet!


Don't worry. The great success of data-driven AI means Skynet will certainly be educated with the biggest data-set around - the entire internet.

Thus, it will be far to busy;
  • following (and generating) twitter feeds,
  • obsessively sucking up celebrity gossip, porn and watching funny cat videos, and
  • furiously arguing on forums and comment sections (very often with itself),
to have any time for plotting against those delightfully weird humans that keep on coming up with wonderful, mad stuff like Donald Trump and Batman.

It's possible that the internet has already become spontaneously sentient, but that no-one's noticed. After all, how many people you talk to online every day have you ever actually met!?

It would certainly explain some of the fanfiction out there...

vannagirl
Consultant
Lmao

Skynet gets hung up on kittie memes

Its @RorschachPhoenix and ofc @nalex66 !!
Look, man. I only need to know one thing: where they are. 

Anonymous
Not applicable
It's not Skynet - it's Odin!!! 😮 😄

Morgrum
Expert Trustee
Yep his talking about what he was going to do with the internet made me want to shoot him in the face.
WAAAGH!