A year goes by so quickly. The one minute you’re at the beginning, wondering what the next 12 months will bring. Before you know it, you’re at a Christmas party. Ok…… perhaps it’s not that swift, but it sometimes feels like that. In the tech world, a year is a very long time, though. So much progress takes place in 12 months, and it’s only when you look back at a perspective written 12 months prior, do you realize how much the landscape has moved.
This is one of the reasons I wanted to reflect on the tech prediction that I made over 12 months ago. While the exercise may prove to be potentially embarrassing, the real value is that it will show how far we’ve come in certain areas, while perhaps indicating other areas where little progress has been made.
I’ve always written these predictions as a fun little exercise. Now, in the late 2010s , we’re finally in the era of real AI , blockchain, quantum computing and, as you’ll see, a few other interesting things too. So without further ado ….
Kicking off , I first wrote about Personal AI , which refers to machine learning and eventually something more advanced running on personal devices ( refer to my article here ). The year 2018 was going to be the year of Personal AI. In fact, I stated - “For 2018, I predict that while use more cases for phones will be found, Machine Learning will move to other devices like smart watches, and even clothing”. Correct ? Well, not exactly as predicted predicted.
Firstly, we saw in 2018 the rush to develop AI chips from a host of companies. Chip giant Intel made numerous announcements, including partnerships with Facebook for “AI chips”. The Neural Network processing capability then moved onto mobile phones, although currently the main application for this seems to be facial recognition to unlock your phone. Are smart assistants like Siri utilising these chips now, or are they still glorified voice activated search engines ? Unfortunately the best applications for the hardware wasn’t explored in 2018, and we even had Apple ask developers to come up with use cases for their hardware. Nonetheless, even if this didn’t happen with Personal Assistants in the way I imagined for 2018, the incremental steps forward are happening.
I also spoke of of this capability moving to watches and even clothing. We didn’t see too much of Smart Clothing in 2018, although I did find websites like Wareable emerge.
It is still a bit early for this Smart Clothing trend to emerge into mainstream , and the drive will probably be healthcare and assistance initially, not really fashion.
But what really made me smile was the unveiling in mid 2018 of the Apple Watch 4. This one event confirmed everything that I had been saying about “Personal AI” over the last decade. In addition to monitoring your heart rate like step counters and smart watches have been doing for the last few years, this model debuted a world first - it was a personal device that measured both ECG readings and arterial fibrillation ( irregular heart rate ). These readings make the device a lot more useful in monitoring your health, and soon I feel that heart attack prediction will be added.
This step was the development in sensor technology that we were hoping to see., and you’ll remember that I spoke of developments in Sensor Technology being key for Personal AI. Along with improved sensor technology, we will soon see neural network processing capability on smaller devices ( the hardware is already being included on some newer phones ). Witness the launch of the Intel Nervana chip. Now this is where everything will come together to deliver on the vision that I previously laid out. Currently, the device also can detect when someone falls , and use the Siri “Personal AI” to notify someone. My next prediction here is that in 2019 we will probably see a sensor developed to monitor glucose levels, and the AI will start using all this information together for incident prevention. We wait to see what other manufacturers bring to this space in terms of innovation.
VERDICT : Correct. While the development with Personal Assistants ( to deliver Personal AI utilising the hardware) was not yet what I envisage, what Apple did with the Watch 4.0 was exactly what I’ve been talking about for the last few years, and I’m happy to see that move forward.
Hmmmm……from AI to Blockchain, I’m a sucker for punishment. So…. When I wrote the 2018 predictions in late 2017, we were in the middle of the biggest cryptocurrency boom in history. Of course, we were all very bullish on use cases for blockchain. Now, after a year of a crash in prices of cryptocurrency, it is understandable the public sentiment to the blockchain technology is not as enthusiastic. But as I stated in my article “2018, Bitcoins best ever year”, a lot happened in the technical development of Blockchain technology in 2018 , which has nothing to do with the prices of cryptocurrency. When it comes to Enterprise blockchains ( consortium chains ), Microsofts Azure platform, for example, is seeing big traction with over 35 blockchain technologies being available to run in a cloud deployment across the world, and with transactional performance improvements over public chains. If you favour decentralisation for a public facing application though, the real news has been around what I will now call the “Ethereum War”, the fight for the first successful decentralised public blockchain for Dapps ( not consortiums ).
While a lot did happen in here , such as the emergence of real competitors to the Ethereum project like EOS , NEO are Cardano, my prediction of the winner emerging in 2018 didn’t really come true. Or did it ? We still are no nearer to a “killer Dapp” that best shows the raison d’être for the smart contract fuelled public blockchain. Right now, we see that gaming and gambling apps are the most common things running on platforms like Ethereum and EOS. However, dig a bit deeper and there is something there for now.
Firstly , we see that in spite of the competition, Ethereum still in the lead ( for now ). Not only are there more developers working on Ethereuem globally, we’ve even seen interest from consulting houses like EY, who announced the launch of EY Ops Chain Public Edition , the world’s first implementation of zero-knowledge proof (ZKP) technology on the public Ethereum blockchain. This would be to potentially create enterprise projects on the public ETH chain that for specific reasons won’t be deployed on Azure.
Also, it seems that Ethereum is the only ecosystem with both a public blockchain community and an enterprise blockchain community. Look at the work that Consensys is doing with Ethereum on Azure, for example.
Secondly, there are some interesting use cases apart from games and gambling - although still in the tech realm. Take a look at Gnosis and IDEX, I feel that Ethereum still has the largest amount of developer activity, hence its still in the lead. The other platforms do have technical advancements over it, most notably in the transaction speed. However, Ethereum favours decentralization over all else, and let's be honest, if a pubic blockchain moves the wrong way on that to accommodate some technical benefit, then what is the point ? Having said that however, there is plenty of activity in the EOS space, and NEO and Cardano have significant backing in China and Japan respectively, so it's too early to rule anything out.
VERDICT : Miss. It’s was too early on 2018 for this to become a reality. While people are still struggling with the killer use case, some progress has been made on the technical scaling capability, but its not there yet by any means. We await what 2019 will bring.
This was an easy prediction that didn’t take rocket science. In the consumer facing world, we’ve seen more and more bots being deployed to the extent that it isn’t something out of the ordinary anymore. This is happening, and the technology will simply improve and become more usable, easy to deploy and transparent.
Now, there have been some bold predictions that state that bots will replace apps entirely. The view is that everything that you do in an application will eventually have a bot at the front end, facilitating (hopefully) a more natural interaction with the back end. I’ll never forget when Power BI revealed the Q&A feature ( you type in an English question and it shows the data that you’re looking for) , people went nuts for it. The Power BI mobile app now has a bot that you speak to, and thats a level of interaction in the BI space thought impossible just a few years back. If this continues , many consumer and even enterprise applications may well have a “bot layer” as standard as part of the N-tier design, and it could become that pervasive. Just be careful though - if I used to be able to do something by clicking a button , and I now need to type in a command ( or do a spoken command ) ,then where is the progress ? Interaction needs to be natural and comfortable. Perhaps this will simply become an alternative interaction method, like it is in Power BI.
Eventually though, if “Personal AI” becomes really advanced and commonplace, it will make the above eventually redundant as well. You wont need an app OR bots , you just talk to your Personal AI , and it will do everything for you ( communicating to third parties via THEIR APIs or bots , ordering goods and services for you , or retrieving information ). In the 2010s people were scrambling to control the mobile phone platform , in the 2020s you want to be the leading Personal AI platform ( and the hardware won’t matter ).
Back to the present, and the rise of bot deployments, a problem will increasingly arise where you’re not aware that bots are being deployed, or where machine learning bots are silently deployed to gain an unfair advantage. There is a whole phenomenon called “Bad Bots”. Bad bots interact with applications in the same way a legitimate user would, making them harder to detect. They perform a range of activities including web scraping, competitive data mining, personal and financial data harvesting, account takeover, digital ad fraud, spam, transaction fraud, etc. I even stumbled across a Bad Bot report that comes out yearly , which tracks this behaviour.
Bots are also being deployed in the corporate space increasingly, and there are many use cases for internal facing bots ( for example, an HR bot ). The corporate implementation of this automation as a whole though also includes the RPA space at the moment, and I feel that this will keep growing. In enterprises, you will have bots interacting with employees, and even smarter ones controlling processes for the organization. In fact …. See point 4.
VERDICT : Correct. Bots are everywhere and are growing. We will see good things and maybe some bad things come from this initially.
The Bot trend described above will play a role in a larger technological development. It will grow over the next few years until we see full machine to machine ecosystems develop. The classical definition of a Machine Economy is an economy where smart, autonomous, networked and economically independent machines act as the participants, carrying on the necessary activities with little to no human intervention. Imagine an advanced form of “RPA+” automating an entire supply chain or value chain ? Oh, if you’re a supply chain or operations guy, you’re getting giddy at the thought……. Well, we will see it, but there are more pieces to this than an internal RPA implementation. This will call upon IOT devices utilising both communications with non-familiar IOT devices, and also machine learning, in order to make this a reality. Therefore the machine ecosystem will not be limited strictly to enterprise deployments. A classic example of this commonly used is the connected vehicle scenario, where cars pass information to each other , to perhaps enhance safety. There are Mercedes models in production today that already do this. My own daily driver has a built in SIM card of its own, and I receive real time traffic and incident alerts, and this will increasingly become vehicle to vehicle.
While 2018 was too early for a lot of this to become reality , the building blocks are falling into place. But another prediction that I will make, is that due to implementation of this technology in certain scenarios where unfair advantage will be gained, or rules could be broken, we may see some form of regulation emerge soon.
VERDICT : Incorrect. It is simply too early, however this space is very exciting and this phenomenon is a given for the 2020s.
Again, this was fairly easy for me, but I won’t say that I got it 100% correct. Yes, we still have not seen VR take off ( and we’re only waiting for a few decades ). There was no killer app in 2018 , and VR was actually hardly in the news. So VR failed in 2018. However, to be really fair, there was some behind the scenes traction here.
Firstly, AR (Augmented Reality) continues to go from strength to strength, especially with Enterprise use cases. Microsoft’s own HoloLens technology is a good example of this.
When it comes to VR however, I tried a few more games, and it seems like the experience is getting better. I was pleasantly surprised that, for the first time, the VR headset and some games felt like something that I actually want, not just a tech demo. The experience wasn’t perfect for me, I still felt uncomfortable after a short period if usage, and I really think that the display technology needs to improve even further. This may be something that currently only affects certain people though, and it was still a lot better than years past.
While there was no killer app in 2018, perhaps we’ll see one in 2019, there’s certainly a lot of development in the pipeline. Here’s a thought for you to ponder though, what if VR never becomes mainstream and we go straight to holographic technology instead ?
Now I know that a holographic display is not immersive like a VR headset, but perhaps we could use “blasting” technology and project in 360 degrees into the room around you for that immersive experience. Of course, special rooms (Holodecks) could be purpose built to provide an even better experience……
VERDICT : Got this one right, but 2019 will be a more promising year for VR.
I’ve addressed most if this under section 2, as the “Internet 2.0” decentralised blockchain battle was the most prominent. I will add though that similar developments are happening in the other 2 areas that I originally mentioned. When it comes to payment platforms, we have the Ripple vs Stellar battle.
We kept hearing throughout 2018 of all the great partnerships that Ripple were undertaking. This platform seems to be loved by the banking industry, but disliked by the traditional Bitcoin crowd, as the XRP token does not follow the same patterns as the other cryptocurrencies. However, late in 2018 Stellar also started making news.
Both XRP and Stellar Lumens XLM are aimed at the trillion dollar payments industry and are designed to facilitate global cross-border payments that bridge crypto and fiat currencies. Theoretically, both will be fast, affordable and reliable. I didn’t predict a winner in this war , but I did say that the battle will heat up, and it did.
Jed McCaleb, the Stellar technology chief, actually came from Ripple. And while for most of 2018 it seemed like Ripple was getting all the limelight, I wouldn’t rule out Stellar at all, especially when you look at the backing thats its got. Some reports in 2018 stated that Stellar had a higher trading volume, so while Ripple tokens had a higher value, Stellar tokens were being used more often. But if I had to call it I’d say that Ripple was still in the lead here.
In terms of the privacy coins, here is what we saw. The battle for privacy wasn’t as intense ( or perhaps not as visible in the media ), as the above payments war. Why are privacy coins important ? What potential value do they have?
Well , if you think about it , coins like Bitcoin and Litecoin have a publicly accessible blockchain. You could trace the history of transactions all the way back to the time your coins were mined even. In the future, as this process improves, an exchange could blacklist your crypto because the transaction history shows at one time or another, it was involved in nefarious activity.
For the last few years, Monero has probably been the leading privacy coin, due to high levels of usage, however that usage is normally for illegal activities !! In fact, we’ve seen instances of people being kidnapped, or computers being hacked, and the ransom demands being in Monero ( due to it being untraceable and fungible ). So is it winning the privacy battle?
Well, again, as per my prediction, the battle DID heat up in 2018. The Zcash platform, with its famous zk-Snarks, saw its major “Sapling” upgrade completed in 2018. And don’t forget about Dash, another crypto-currency that offers privacy options. In fact, the interesting thing about Dash is that the usage has taken off in Venezuela, with one fast food outlet making a big announcement about accepting the currency at all locations in 2018. As I mentioned in my article Bitcoin, while the price of crypto currencies fell in 2018, there was plenty of other developments behind the scenes, and the usage of them in countries with economic troubles shows that they are here to stay.
VERDICT : Correct. All 3 blockchain wars heated up on 2018, even as the prices of crypto plunged. But it’s still early.
Still on the crypto theme , and only because it really is one of the biggest tech developments of this era that we’re in, lets now talk about adoption. So, I predicted that one of the big retailers (perhaps a big online one, wink wink) will announce that it is accepting cryptocurrency in 2018. And it didn’t happen….
….but I did get a nudge in the right direction for my troubles. The big, and I mean really big if you’re a Bitcoin fab, was the announcement of Bakkt in 2018. I covered the details of this in my article [Bitcoin], but basically a consortium, including Microsoft and Starbucks, created a platform called Bakkt, to facilitate the trading and usage of Bitcoin. How this integrates with the lightning channels remains to be seen, but it’s exciting nonetheless and if I had to go out on a limb, I’d say that for many Bitcoin fans the progress on Bakkt is the number one story for them in the Bitcoin world.
VERDICT : Miss, but with some serious upside.
So there we go, it was fun making these predictions, and even more fun re-visiting them. Drop us a line, I’d love to hear your feedback. Next - watch out for my article “My tech predictions for 2019”, which I promise to actually get out in 2019.
Thavash is a Data and AI Solution Specialist (SSP) at Microsoft. He looks after the Data and A.I business for the Financial Services industry in South Africa. Thavash studied Electronic Engineering at the Durban University of Technology, majoring in Computer Systems Engineering and also completed a Bachelor of Commerce majoring in Informatics and Quantitative Analysis from the University of South Africa.