AI / Robot Child Care in ’50s Japan

It sort of feels like the 1980/90s to me, or for my parent’s generation the 1950/60s. With all the marketing jabber on artificial intelligence, robots / drones, home / city automation, self-driving or flying cars, virtual reality, ‘modern’ homes, the difference today is its mashed-up up with prepositions such as cyber, digital, smart, and personal. Its an industry in a peculiar state.

I’ve pointed out the silliness of the term digital for years. Same with the pejorative term ‘over the top’ (OTT). So now onto ‘Cyber’, it means nothing – and that’s from its creator. Cyberspace comes from the science fiction writing of William Gibson, particularly his 1984 novel Neuromancer. Gibson’s account of how he coined the term cyberspace was he needed a “really hot name” for the arena in which his stories would be set, and cyberspace “sounded like it meant something or it might mean something, but as I stared at it, my whole delight was that I knew it meant absolutely nothing”.

Similarly ‘disruption’ is reaching the same level of nonsense in it means nothing, change happens. Just because the rate of change and its reporting is a little more frenetic doesn’t make it unique. Some things change, the world generally remains the same. Some businesses can not adapt to specific changes. Generally they see the change coming, like Kodak, just are unable to help themselves, as business as usual kills innovation.

On AI, changing a value in a database because of a difference in average measured state is not artificial intelligence as portrayed in the current crazy visions. Google knows where my home and work are through a simple algorithm (e.g. middle of night and location same as most other nights, then location=home). Same as telling you at 5:30 PM on a Tuesday how long it will take to get home as that is when you usually head home on a Tuesday. We are creatures of habit. This is not some awe-inspiring herald of AI. In the past 20/30 years we’ve come a long way from neural network research, and from the 50/60s with fuzzy logic. But we have a long way to go.

At this point ‘corporate astrologers’, so-called ‘futurists’ will go, “You know nothing, Alan Quayle!” Peddling fear through unforeseen dramatic change is their business. Telcos clearly see their demise into ISPs, its just whether they can help themselves. Corporate astrologers are not going to help them with that.

AI will continue its rise, but in small incremental ways, for example in BOTs (mainly text based ones) across many vertical applications. We’ve talked about such bots like KISST before and many hacks using them at TADHack.

On the connected home, finally reality is starting to set in. Nest and Dropcam are allegedly struggling. I’ve used many WiFi cameras over the years, I reviewed Home Camera back in 2008, its fun, but the use case becomes less compelling as situations change, e.g. helpless babies become toddlers etc. I have 5 thermostats around the home, and all are set after way too many hot/cold battles to a program that is acceptable to all. I can not afford for the sake of family harmony a device making decisions to lower temperatures because everyone left the house on a cold day for someone to come back to a cold home. But that’s just a personal example which is irrelevant. In hot countries like Singapore, within a few minutes of turning the AC on the place is cool, a networked thermostat has little benefit to most people. On the connected fridge, here’s a simple solution, put a bloody whiteboard on the fridge and list stuff out when its getting low, then whoever is going shopping take a photo if their memory is poor.

All uni-tasker WiFi or bluetooth devices around the home only need to work securely (and privately) between the app on one of my computers (of any form factor) and those uni-tasker devices. No third party spying (data gathers), thank you. The same is true in most industrial applications where security is paramount. Now in smart cities, its a little different – openness in the data is important, BUT security to the devices remains critical. Its the Internet of Silos!

Virtually Reality still hurts your eyes after 5-20 minutes depending on age, the same was true in 80/90s VR headsets. Same is true with 3D TV today. When will we ever learn?

The end result of all these moon-shot visions is when small, simple, incremental services are shown, they are treated with comments like, “that’s just too small for us to bother with”, “where’s the business case”, “we’ve never made money selling VAS to telcos”, “telcos can’t sell VAS.” Yet all those small services are easy to understand, and solve compelling problems for which people are prepared to pay. Slack is just a very simple enterprise messaging platform, that forces you to follow it regularly else loose track of the conversation. But look at its crazy valuation and the massive ecosystem its building.

The industry is in a peculiar state, stop acting like sheep for fear of loosing your job, express your well-founded incredulity at the relevance of the MWC VR demo, and at so much of the tripe being pedaled today in corporate slideware and websites. Its happened many times before, history is repeating itself. Get focused on the small incremental stuff that moves us forward. Chasing butterflies will just repeat the errors of waiting for the big IMS/RCS future while other IP-based communication services have filled the many small gaps in voice, messaging, video, security, and many specific use cases and integrations.  Small stuff rocks!