Facebook F8 – Business as usual

Reply to this post

RFM AvatarSmall

 

 

 

 

 

Progression, but little excitement.

  • While Facebook’s announcements at its developer conference were not exactly ground-breaking, it is clear that Facebook remains extremely commercial and is very focused on fixing its weaknesses.
  • Highlights from F8 include:
    • First augmented reality: Facebook has decided to re-focus on AR in what looks like a pre-emptive strike against Snapchat.
    • This includes the launch of a new camera platform that follows up on the roll-out of camera capability to all of its apps.
    • With the camera capability now everywhere, developers will now be able to create content that can overlay the real world as seen through a smartphone camera.
    • This includes Snapchat-like photo and video annotations as well as combining the ability to map a 3D environment and place virtual objects within it.
    • The aim here is to get users to spend more time within the Facebook ecosystem thereby increasing potential for monetisation.
    • Second: Artificial Intelligence. This remains a major weakness for Facebook but it does appear to have made some progress in image recognition.
    • This makes some sense as the core competencies of its biggest AI hires are in this area.
    • Facebook showed AI that was capable of advanced object recognition as well as the tracking of moving objects through video.
    • This was complimented by progress on making Facebook M (digital assistant) smarter as well as improving the quality of the bots that it offers to companies to communicate with their clients.
    • AI remains essential to Facebook’s long-term growth as it is sitting upon a mountain of data but still is not in a position to really make the most of the insights and automation that it can provide.
    • Third: Gaming. A lot of progress has been made in developing game play within Messenger.
    • Rich game play is now enabled with real time and turn by turn games making up the majority of the line-up.
    • There are now 45 games available for play within Messenger and with the gaming tab be enabled within the app imminently.
    • I think that this is a crucial step forward as gaming remains the largest segment of Digital Life and in developed markets, there is still no dominant player.
    • Fourth Chat extensions: This enables developers to bring their services directly into chat sessions.
    • Spotify is the best example where users can search Spotify for a track and then post that track as well as play a sample all within the chat.
    • Apple Music will also be coming to the platform later in 2017.
    • Fifth Monetisation: Behind all of the new announcements is a single-minded determination to drive more traffic onto the platform.
    • This can be seen everywhere from the desire to enrich mundane events to encourage sharing to enticing users to spend more time on the platform.
    • Data richness and time spent are two key elements when it comes to understanding user activity and being able to earn revenues from it.
    • This is not a subject that was directly addressed anywhere but one can see the hand of the cash register in everything that Facebook does.
  • Facebook revealed nothing that was really ground breaking but instead spent time addressing its weaknesses as well as ensuring that rivals are not able to steal its user engagement.
  • I am still quite cautious with regards to Facebook’s outlook for this year as I don’t think that either its video offering or its gaming offering are mature enough to bring the company back to high growth in 2017.
  • This combined with requirement to really improve its AI to compete on a level playing field with Google, Microsoft, Apple and Amazon leads me to still prefer Baidu, Tencnet and Microsoft.

 

Google – From Russia with love pt. III

Reply to this post

RFM AvatarSmall

 

 

 

 

 

Google escapes to fight another day.

  • Google has settled its dispute with the Russian regulator but I suspect that Google has managed to avoid having to give away its crown jewel: Google Play.
  • According to the FAS Russia (see here), Google has agreed to:
    • First: Google will no longer prevent the pre-installation of competing search engines or apps on Russian GMS-compliant Android devices or their presence on the home screen.
    • There is no mention of which services will be set as default when the user turns on the phone for the first time.
    • Second: Google will no longer enforce the parts of the previously signed agreements that contradict the terms of the settlement.
    • As the terms of the settlement have not been made public, it is not clear exactly what this entails but I suspect that it refers primarily to point 1 above.
    • Third: Google will ensure that users of Android devices already present in the market will be given the option to change their default search provider via a software update and pop up.
    • Fourth: Google will pay a fine of $7.8m which I calculate is equivalent to around 100 minutes of Alphabet’s cash flow and consequently, is completely irrelevant.
  • This settlement ensures that competing apps can be on the home screen but it appears to do nothing about the requirement to bundle the Google Apps with Google Play nor the fact that they are set by default, albeit, now changeable.
  • The settlement was proposed by Google and accepted by the FAS which admitted that it was under some pressure to have this two-year dispute resolved.
  • This is why I believe that this crucial element was left out if the agreement as I think that it is the unbundling of Google Play from Google’s Digital Life services that could do the real damage.
  • This is because it is widely accepted that in most markets outside of China, it is almost impossible to sell an Android Device that does not have Google Play installed because this is what users demand.
  • This gives Google the power to force handset makers and operators to install the services from which it makes almost all of its Android mobile revenues front and centre on the device and to set them as default.
  • Research has shown many times that installation at the factory and being set as default are big drivers of usage, even if the service in question is considered to be inferior (Apple Maps).
  • This is why I have long believed that Google Play is so important to Google’s Android revenues and that unbundling Google Play could be highly detrimental to its long-term outlook.
  • The good news for Google is that it has now set a precedent with which it may be able to more easily settle its outstanding dispute with the EU.
  • The net result is that I think that while Google has given up a little ground, the fortress of Google Play remains intact and with it its ability to continue dominating the Android landscape.
  • However, this dominance is not enough to make me think that the shares are attractive and I continue to prefer the shares of Baidu, Tencent and Microsoft.

India e-commerce – The big if.

Reply to this post

RFM AvatarSmall

 

 

 

 

 

Flipkart for Snapdeal looks increasingly likely.

  • The probability of consolidation in Indian e-commerce is creeping ever closer as Softbank, is pushing for the sale of Snapdeal to Flipkart at a valuation considerably less than the $6.5bn at which the company last raised money.
  • I think this move makes complete sense as on their own, both Flipkart and Snapdeal are likely to be crushed by Amazon should it decide to pull out all the stops in order to dominate the Indian market.
  • This is because, on their own, neither of them is large enough to keep Amazon at bay, but together, they might just have a chance.
  • Snapdeal and Flipkart like Alibaba and to a lesser degree Amazon are market places which bring together merchants and buyers in one easy to use location and from which they can take a small cut.
  • In effect, they are network businesses just like Uber, Alibaba, AirBnB, Craigslist and so on and consequently, they are bound by the same rules.
  • 18 months ago I proposed a rule of thumb that states: A company that relies on the network must have at least 60% market share or be at least double the size of its nearest rival to begin really making profit (see here).
  • This, in a nutshell, is the problem faced by both Flipkart and Snapdeal in India.
  • Flipkart is bigger than Snapdeal and so it is in a slightly better position but it is not double the size of its nearest rival.
  • Data from 7ParkData shows that Flipkart has about 35% of e-commerce monthly active users followed by Amazon at 23% and Snapdeal at 13%.
  • As it stands today, not one of the Indian e-commerce players has established itself as the go to place to buy and sell goods, meaning that all parties are likely to be losing large amounts of money through aggressive competition.
  • If Flipkart is able to successfully execute the acquisition of Snapdeal and hold onto all of its users, then its share of MaU will reach 49% more than double that of Amazon.
  • This could give it just enough scale and momentum to become the go to marketplace in India making extremely difficult for Amazon to compete.
  • This is a big if and will require flawless integration, streamlining as well as customer service.
  • This is why I suspect Softbank is willing to take a substantial haircut on its investment as I think it has concluded that should Snapdeal remain independent, its investment could easily be worth nothing.
  • Amazon does not have a good track record in emerging markets as its performance in China vs. Alibaba was dismal and it does not seem to do much with its acquisitions other then leave them to their own devices.
  • Hence, I think the combination of Flipkart and Snapdeal has a chance but Amazon does seem to be determined not to repeat in India the mess it made in China.
  • Valuations are falling, highlighting the prospect of bargain hunting, but the high-level of uncertainty keeps me from wanting to be involved.

Spotify – Crown jewels

Reply to this post

RFM AvatarSmall

 

 

 

 

 

Spotify keeps the crown jewels to itself.

  • In striking a deal with Universal, Spotify has traded much better than I thought it would giving the label two concessions that I think will end up being pretty worthless.
  • Spotify has signed a licence with Universal that has three main aspects:
    • First: Universal artists will have the option to release their music to premium users only for two weeks before it is available to all users.
    • Second: It looks like Spotify will cut the share that pays to Universal from 55% to 52%.
    • Third: Spotify will provide Universal with the data that its music generates thereby enabling the label and its artists to gain better insights into how its music is being consumed.
  • On the surface, it looks like two of these points benefit the labels but when I take into consideration how the music industry is evolving, I think the winner from this deal is Spotify.
  • This is because Spotify has managed to increase its gross margins on Universal music by 300bp and has cleared one major hurdle towards its road to an IPO.
  • Sony and Warner are the two remaining hurdles which, now that a precedent has been set, may be easier to overcome.
  • That is what Spotify has gained from this deal but what has it given up in return?
  • Not much in my opinion.
    • First: I do not think that delaying releases to the free tier for two weeks will have much, if any, impact on the appeal of the free tier.
    • I have long believed that the free tier is far more valuable to Spotify than anyone thinks that it is (see here) and I think that its desire to protect the user experience of this segment has been a major sticking point in striking a new deal with the labels.
    • Time shifting media releases is how music and films have been released for years but I think that this is changing.
    • Spotify knows what it users listen to and what they like and I think that in the future, users will be increasingly made aware of new music when the streaming services recommend them rather than when the artist or label releases them.
    • Hence, many users might not even notice a two-week delay meaning that Spotify has received something for nothing.
    • Second: Spotify is giving Universal access to its fire hose of data but I have doubts whether it will be able to make much sense of it.
    • This is because it will only have access to the data which is just a raw material.
    • To make something useful out of it, trained algorithms are required to parse that data and draw meaningful conclusions from it.
    • These algorithms are Spotify’s crown jewels and I am pretty sure that they will be staying safely under lock and key.
    • After all, they are the reason why Spotify’s service is better than Apple’s and are the key to its ability to eventually replace the labels entirely (see here).
    • Furthermore, Universal will only have access to its own data which compared to the entire catalog that Spotify has, is a small subset.
    • Hence, even if it could make sense of the data, it wont be able to draw many meaningful conclusions from it as it will be looking at an incomplete picture of user activity.
  • The net result is that I think Spotify has dealt much better than I thought it would as I was concerned that the pressure to make it to IPO in 2017 would force it to give too much away to the labels (see here).
  • In fact, I think that the reverse has happened putting Spotify in a good position to IPO without having to give much, if anything, away.

 

Twitter – End of days

Reply to this post

RFM AvatarSmall

 

 

 

 

 

The loss of the NFL is a disaster.

  • Twitter has lost its deal to stream certain NFL games which I think punches a potentially fatal hole in its strategy to break out of its niche of microblogging.
  • Amazon has reportedly paid $50m for the rights to stream 10 Thursday night NFL games for the coming season.
  • This is 5x what Twitter paid in 2016 but I do not think that this is a case of the little guy being priced out of the market.
  • Instead, I think that Twitter got a very good price from the NFL because of its promises to be able to leverage its social interest graph to generate meaningful advertising revenues as well as insights that could be shared back to the NFL.
  • Clearly, Twitter has not been able to live up those promises which is why the rights have been sold to a more conventional bidder who I think is simply paying a more regular price for the rights.
  • In my opinion this is nothing short of a complete disaster because expanding into media consumption was Twitter’s one hope to break out of its niche and resume subscriber and revenue growth.
  • The loss of the NFL is an indication that this strategy is failing and that despite its efforts, it is nothing more than a broadcaster of short text messages and a second-rate instant messaging platform.
  • Blogging and Instant Messaging make up a total of 16% of the Digital Life pie which I have long believed that Twitter has already fully monetised.
  • I remain convinced that this is the reason for its growth grinding to a halt (see here).
  • If Twitter can entice its 300m users to do more with Twitter beyond these activities, then there is scope for revenues to begin growing again as it will have more traffic to monetise.
  • This is why the video strategy was so important.
  • Media Consumption makes up another 10% of the Digital Life pie and had Twitter been able generate significant traction from it, there would have been significant upside from current revenue levels.
  • Without this growth, I still fear that Twitter’s shares will fall below $10 because even at $14.5, with no growth, the shares are still expensive.
  • This loss makes it even more likely that 2017 is going to be a stagnant year where the realities of the company’s predicament really begin to become apparent.
  • I think that this could drive the shares to $10 or below.
  • I continue to see Twitter as a potential acquisition target but would expect to see the shares touch $10 before real interest is triggered.
  • I see no reason whatsoever to go bargain hunting as there is no bargain to be had.

Google – Meaningless milestone

Reply to this post

RFM AvatarSmall

 

 

 

 

 

Android’s success shows up Google’s deficiencies.

  • Android has surpassed Windows as the No.1 platform for accessing the Internet globally, highlighting just how bad Google is at monetising Android as it remains only a small percentage of total revenues.
  • I think that this could be a growth opportunity if Google can fix the many problems that exist within the system that it created and in many cases controls.
  • According to StatCounter, Android devices now make up 37.93% of all Internet access devices very slightly ahead of Microsoft Windows at 37.91% with iOS a distant third at around 13%.
  • Furthermore, with most users spending more time on smartphones and tablets than PCs, it is clear that the PC is rapidly becoming a device used in the enterprise and by content creators.
  • This is a major reason why RFM does not consider PC usage as a contributor to Digital Life when assessing the addressable market for a digital consumer ecosystem.
  • Consequently, it would be natural to assume that Android is a big part of Google’s revenues but in reality, it is not.
  • RFM estimates that in 2016 just 19% of advertising revenues came from Android devices compared to PCs and Macs which generated 60% of advertising revenues.
  • A further 19% of revenues came from iOS devices despite the fact that there are 2.9 Android devices for every 1 iOS device.
  • This tells me that the PC is a much better platform for advertising monetisation but it is also a clear indication that Google is doing something very wrong when it comes to making money from Android.
  • I have long argued that while demographics plays a role, the endemic fragmentation of Android and Google’s inability to update software on its own devices severely hinders the usage of and loyalty to, the Android platform (see here).
  • I believe that this is a major reason why an Android device generates less than half the revenues that an iOS device does which is also meaningfully less than a PC or Mac.
  • While this is a real black eye for Google, I also see it as an opportunity.
  • RFM estimates that in Q4 16A each iOS user delivered $3.37 in revenues for Google compared to $1.47 on Android.
  • If Google could fix the problems with Android, then I think that there could be meaningful upside to this number.
  • For example, if Google was able to increase monetisation of its own Android devices to $2.00 per user per month, this would increase revenues by $6.4bn on an annualised basis.
  • As smartphone user growth and usage both slows, Google will need to look for growth elsewhere and I see this as an obvious place to start.
  • I am hoping to see signs of this at Google i/o (in May) but in the preview of Android O (see here), I was disappointed.
  • Without these kinds of actions I think that Alphabet remains fully valued and would prefer the shares of Microsoft, Tencent and Baidu.

Didi – Nasty economics

Reply to this post

RFM AvatarSmall

 

 

 

 

 

Didi encounters a real growth problem.

  • Didi is now the only credible ride hailing platform left in the Chinese market, but its problems are far from over because local Chinese governments are finally enforcing regulations designed to protect the taxi industry.
  • In December 2016, the governments of Shanghai and Beijing approved a policy called local cars, local drivers.
  • This meant that ride hailing companies could only use drivers who had locally registered vehicles and could prove residence in the city.
  • This may not sound like a big deal until one looks at the demographics of the urbanised workforce in China.
  • Around 40% of the workforce of both of these cities reside outside of the city and in the younger, paid part of the workforce, that number is much higher.
  • For example, prior to the enactment of this regulation, less than 3% of Didi’s Shanghai drivers had the necessary residential registration to qualify as drivers.
  • I suspect that that Didi’s Beijing drivers show similar characteristics and that other major Chinese cities also have a large migrant workforce.
  • This has not become a problem until very recently because although the policy has been approved, it has not been enforced until very recently.
  • The result has been that Didi has now been forced to substantially reduce the number of drivers in Shanghai and as of April 1st, is only allowing cars with a Beijing licence plate to operate in Beijing.
  • The net result is that supply of rides has been drastically reduced meaning that the price of those rides will inevitably increase.
  • Furthermore, on top of the service becoming more expensive, the quality of the service is also likely to meaningfully decline.
  • With less cars available, wait times will certainly rise and the chance of successfully hailing a ride will decrease.
  • Rising prices and lower reliability is likely to drive many users back into the arms of the taxi industry thereby achieving exactly the result for which the rules were created.
  • Beijing and Shanghai are the two cities which have the largest non-resident workforce but I suspect that this sort of legislation could easily be used in many of the other large Chinese cities.
  • This creates a very serious problem for Didi as, with its supply of drivers substantially limited in the areas where it has the highest demand, there will be a real crimp on its ability to grow.
  • Furthermore, there remains the very real risk that other major cities in China follow Beijing and Shanghai’s lead causing Didi further agony.
  • If this spreads, it is not inconceivable that Didi’s revenues start going backwards.
  • Didi was originally created as a taxi booking service and one possibility is for Didi to return to its roots.
  • The other is for Didi to move into the high end (where Uber started) and develop a black car offering but I think that demand for this will be quite limited.
  • This leaves Didi’s outlook at the mercy of regulation, which is not what one wants to hear when one has put money into the company at a valuation of $34bn.
  • DJI or Ant Financial are the only two private Chinese unicorns that I would be willing to consider.
  • In the listed sphere I still prefer Baidu and Tencent.

Tencent – Home circuit.

Reply to this post

RFM AvatarSmall

 

 

 

 

 

I think investing in Tesla is about China, not overseas.

  • Tesla has announced that Tencent is investing $1.8bn in a 5% stake in the company that should bring in both badly needed capital as well as serving as a way to open doors in China.
  • Some commentators are of the opinion that this investment is part of Tencent’s plan to expand overseas, but I see this as a move to catch up with what is going on at home.
  • This fits nicely with Tesla’s ambitions in China which has had a pretty torrid time there since it launched in 2014.
  • RFM research clearly shows (see here) that Tencent is the dominant ecosystem in China with over 850m MaU and 77% coverage of the Digital Life Pie.
  • However, when it comes to automotive, I see it as being behind both of its major domestic rivals, Alibaba and Baidu.
  • Baidu has leadership in maps, leadership in autonomous driving and is already putting its services into cars with Baidu CarLife in conjunction with Harman.
  • Alibaba is also actively pursuing Chinese car makers to use its Yun OS software in their cars and already has a deal with SAIC.
  • On the other hand, Tencent has very little other than a small stake in Didi which is unlikely to help it much as both of the other BATmen are also shareholders.
  • This is why doing a deal with Tesla makes some strategic sense on both sides of the table.
  • Tesla gets a strong partner to help it fix the problems that it has had in terms of penetrating the Chinese market, while Tencent gets a potential route into both the automobile and autonomous driving with a major global player.
  • I view the automobile as another device alongside the smartphone, tablet, console, TV and so on for delivering Digital Life services to users.
  • Chinese cities have some of the longest commuting distances in the world (see here) meaning that the automobile is a place where those that drive spend a meaningful proportion of their day.
  • Much of the commuting is done on public transport which Tencent already has covered with its very strong presence on the smartphone.
  • Tencent will be a little more than a passive investor, being referred to as an advisor, which I take to mean helping out in China.
  • This is the latest of a series of moves that has seen Tencent address some of its weaknesses (see here and here and here) but the biggest one still remains.
  • I have long believed that integration is key to monetisation and see substantial upside in Tencent’s financials if it really embraces this notion.
  • However, I still see no evidence that Tencent is making any moves to integrate its Digital Life services into a single system where it can really understand the profiles of its users across all of its services.
  • As a result, its growth is likely to continue to come from selling games and media rather than from its ecosystem which I think limits its real long term upside.
  • Fortunately, the short-term picture remains pretty healthy and so there is still upside to be had without Tencent really developing its ecosystem, but the time is fast approaching when it will need to make that move.

Google & Facebook – Trouble in paradise.

Reply to this post

RFM AvatarSmall

 

 

 

 

 

Facebook will be hit much harder than Google.

  • Although it is Google that is taking most of the heat in the current boycott of its digital display advertising, it is Facebook that is likely to be hit hardest by this problem.
  • A series of global firms have pulled their advertising from YouTube is response to finding their advertisement placed in videos which are deemed to be offensive or contain extremist content.
  • This is not a new problem but has come to a head after a series of lapses on the part of YouTube have gone viral raising the ire of companies who would appear to endorsing such content.
  • The net result is that a number of multinational companies (more likely to follow) have pulled their advertising from YouTube until they are confident that Google is able to ensure that their logos and advertising only appears along aside acceptable content.
  • This is a difficult problem because YouTube is adding 400 hours of content to its website every minute and thousands of websites are added to its network on a daily basis.
  • This makes monitoring content on a proactive basis extremely difficult which is why a meaningful number of lapses have come to light.
  • Fortunately for Google, I think that the impact of this issue will be limited and short-lived as:
    • First: this issue only affects YouTube and display advertising which RFM estimates makes up just 12% of gross revenues.
    • Hence, even if this were to fall to zero, the vast majority of Google’s business would be unaffected.
    • Second: Google is the best equipped to deal with this problem compared to any of its competitors.
    • The amount of content that has to be checked on a daily basis is so vast that it can only realistically be carried out by a machine.
    • This means that AI is needed to scan uploaded content and new websites and flag any that are suspected to contain content that Google customers are likely to consider objectionable.
    • Google has the best AI available when it comes to image and video recognition as well as sentence and text recognition.
  • Hence, I think that Google should be able to fix this problem in a comparatively short period of time.
  • However, I do not have the same degree of confidence when it comes to Facebook which already has this problem but has yet to suffer a loss of businesses as a result of it.
  • I think that when this does happen at Facebook, it will be a much more serious problem as this type of advertising is a much larger part of its revenue and I do not think that Facebook has the AI to fix it.
  • RFM research (see here) has found that Facebook is far behind its global peers when it comes to AI, mainly as a result of having not worked in the field for very long.
  • This means that while it is sitting on the world’s second largest treasure trove of data, it is unable to understand what most of it is and is therefore unable to weed out content that is objectionable to its customers.
  • I think that this will take a very long time to fix compared to Google and so when this issue hits Facebook, it will hit it harder and it is likely to last longer.
  • This is just another reason for me to remain pretty cautious on Facebook which I think will see much lower than expected growth this year which is likely to take a toll on the short-term valuation.
  • I also think that Google remains fully valued and would prefer the shares of Microsoft, Tencent, Baidu and Apple for long-term value based investors.

Tencent – Brute force.

Reply to this post

RFM AvatarSmall

 

 

 

 

 

Tencent has time as most gaming AIs are relatively simple.

  • Tencent is finally jumping into artificial intelligence (AI) and I think that it is fortunate in that it is not very difficult to create good algorithms for the vast majority of the games that Tencent offers.
  • Tencent has created AI Lab which now has more than 250 employees whose task it is be to create algorithms that create more sophisticated game opponents as well as chat bots for companies that use WeChat and QQ to talk to their customers.
  • The general impression of AI seems to be that as soon as one has created an AI group, superb, hyper-intelligent algorithms will come rolling out of the door but in reality, this is very far from the truth.
  • RFM research has found (see here) that single biggest determinate of AI excellence to date is time and those that have been doing it the longest tend to have the best AI.
  • This is why RFM has found that it is the search engines thatare the most advanced even though some of the biggest brains in the field are employed elsewhere.
  • With Tencent just getting into this field, I think it will be a very long time before it will be in a position to roll out algorithms that are capable of making its services meaningfully more intelligent.
  • In the long run this will be crucial to maintaining its dominance in the Digital Life segments where it is present as I think competition will become much tougher as the market matures.
  • The good news is that it is unlikely to prove very difficult to create algorithms that are more than good enough to play the games that it offers to a very high standard.
  • This is because most games are either based on hand eye co-ordination or can be solved by an algorithm using a brute-force approach.
  • Brute force involves evaluating every possible outcome from a given position and choosing the best one.
  • With today’s improvements in memory and processing power, this is not very difficult to achieve.
  • The highest profile exception is Go which has so many possible combinations that brute force becomes impossible.
  • This is why DeepMind’s AlphaGo was such a breakthrough, as it uses AI to work out which options should be searched much like a human would.
  • Tencent has produced an AI Go player called Jueyi which has been able to play to a very high standard but I think that the design has been copied from AlphaGo.
  • AI is a co-operative field and DeepMind has published most of its methodology and results for the creation of AlphaGo in the scientific magazine Nature.
  • Consequently, I do not view this as a good example of Tencent coming up with an innovation of its own and I think we will not be seeing hyper-intelligent AIs appearing in Tencent’s services anytime soon.
  • However, I think that Tencent has time as its core markets are still seeing steady growth and it should be reasonably easy to improve the AI opponents in its core segment: gaming.
  • I still like Tencent as there remains substantial upside should it really begin to monetise the ecosystem that it has created but it has a very long way to go before it can be considered a force in AI.