The KIT ─ Knowledge & Information Technology
No. 251 - 1 November 2019
Was this forwarded to you?
In This Issue
Lessons from the Dark Side
Gartner Technology Trends for 2020
Seen Recently
Claude Baudoin

Consulting Services
  • IT Strategy
  • Enterprise Architecture Roadmap
  • Business Process Modeling & Analysis
  • Enterprise Software Selection
  • IT Innovation Briefings
  • IT Due Diligence
  • Executive IT Seminars
  • Cloud Computing
  • Security Maturity
  • Software Process
  • Knowledge Strategy
  • Technical Communities
  • Knowledge Capture
  • Taxonomy development
  • Enterprise Social Media
Contact Us:
cébé IT and Knowledge Management
www.cebe-itkm.com
[email protected]
+1 415 870 ITKM
Twitter: @cbaudoin
Archive:
Previous KIT Issues
Forward this issue to colleagues and friends: use the "forward email" link below at left, rather than "Forward" in your email software, to preserve your privacy, give the recipient more options (their own unsubscribe link, etc.) and to give us better click-through data. Thanks!
Lessons from the Dark Side
A few days ago, we were left powerless and in the dark, literally, in the North San Francisco Bay Area as our reviled and bankrupt utility company cut power to avoid accidentally setting fires -- a goal they actually failed at in spite of the blackouts. And we were the lucky ones, as other people lost their homes (this time around, no one died) in the fires.

What are the lessons people are learning, or should learn, from this experience?
  • Don't back up your files on a local hard drive -- your PC and that disk might burn at the same time if you have to evacuate in a hurry. Instead, use a cloud-based service, which may cost you only about $6 per month and doesn't impose a fixed limit. We use Carbonite and are happy with it, but there are others. An added benefit is that you can restore a file from your PC onto your smartphone, which can be useful in some cases -- including if you are able to recharge and operate your phone but not your PC (e.g., in a car).
  • Those restored files may be read-only on your phone, or hard to synchronize back to your PC once changed (or you will forget that the version on your phone contains changes). For documents you modify very often (say, time sheets), consider storing them in Google Drive or OneDrive, so you can seamlessly edit a single copy from multiple devices.
  • If your company provides services to the public, do not host an information page about your outages in your own data center: it will probably not resist the traffic spikes (or it may run out of backup power). Our utility company's website was mostly useless during the blackout as they failed to understand this. Instead, host something like "mycompany-outage.com" in the cloud, where thousands of hits per minute won't be an issue; make sure key personnel have credentials to log in and update the site -- and that the responsibilities to do so are well defined.
  • If you have a plain old telephone service (POTS) phone line, perhaps for a fax machine (which doesn't like VoIP), then also keep nearby an old telephone that doesn't require electricity to function. You can unplug the cord from the fax and plug it into the phone, and you're likely to be able to place and receive calls even when the power is off.
Gartner's Technology Trends for 2020
Gartner's annual trend report is always awaited with bated breath, even though some of the predictions can be obvious while other ones are unrealistic and don't come to pass (in 1997, an otherwise highly respected Gartner VP of Research told the audience that by now, we would all carry 10 or more communicating devices of one form or another doing different things; she completely missed that smartphones might evolve to perform all those functions within one device).

This year, if you're not paying a six-digit figure to be a Gartner member, you can get a glimpse of their ten-point forecast in this InformationAge summary. And here is the summary of the summary, with some critical notes:
  1. Hyperautomation: robotic process automation on steroids. Since RPA is not widely adopted yet, this prediction may be quite premature.
  2. Multi-experience: human interaction through voice, virtual reality, etc.
  3. Democratization of expertise: this sounds at least in part like crowdsourcing of knowledge, and if you watch Quora, let alone Facebook, or even Wikipedia, you know these systems do not work well unless they are heavily moderated; we think Gartner is being naïve.
  4. Human augmentation: okay, this is a rehash of wearable computing; here, the naïveté comes from the human and societal resistance to us becoming cyborgs.
  5. Transparency and traceability: this is important and overdue, and please let it be applied to fake news first! But this ties nicely with concerns about the explainability of AI, the provenance and pedigree of information, the "software bill of materials" efforts underway at the US Department of Commerce and at OMG, and more.
  6. The empowered edge: that's not new if you follow (and you should) the work of the Industrial Internet Consortium.
  7. Distributed cloud: this could be seen as the merger of cloud computing and grid computing, or the generalization of hybrid clouds. It could also be seen as what the largest cloud providers already do: they don't keep all their resources in a single data center. Nothing new under the cloud?
  8. Autonomous things: the only new concept here is the "swarm" idea -- multiple (possibly many) devices acting in concert. Warning: the military is way ahead of the curve on this one.
  9. Practical blockchain: the applications mentioned here are not new, the real question is whether we are moving from proofs of concept to deployments at scale. The train is leaving the station (we know of real examples in an industry as conservative as oil and gas) and Gartner may just be trying to catch up.
  10. AI Security: we're not sure what's different, in the description provided by InformationAge, between AI and other IT capabilities that should be secured. On the other hand, the use of AI for security (e.g., to automatically learn about normal patterns and therefore detect abnormal ones more reliably) would be interesting and still somewhat new.
Seen Recently...
"Two MIT grads built an algorithm that uses a seven-question quiz to pair people with wine they'll like."
-- Taylor Kirley, referring to this Bright Cellars quiz.
Quoted here because it may be relevant to AI, but mostly because the rest of
the news has been such that no one we follow wrote anything funny all week.
Note: you will have to provide your e-mail address in order to see your results.