Radar trends to watch: January 2020
We note three big things that will shape technology in 2020, and we’re tracking notable developments in open standards and security.
3 thoughts for 2020
I’m kicking things off with three quick thoughts for the start of the new year. These aren’t predictions. Rather, these are thoughts that can’t help but be right, and that take a bigger perspective on what’s happening.
- Whether or not Congress passes legislation on tech companies, discussion of regulation will certainly continue, and it will have an effect on the development of technology. It’s hard to do good work when you’re watching over your shoulder.
- The big story in infrastructure and operations (aside from our new conference) will be learning to put machine learning products into production. ML and AI raise challenges that few ops teams have faced.
- Contrarian as this may seem, we are still in the early days of cloud adoption. There is a lot further to go.
Open standards trends
- Amazon, Apple, and Google are working on open standards for smart homes. If smart homes are going to go anywhere, open standards and interoperability are essential. I’m not big on the whole “smart home” concept, and it will be hard for this market to escape suspicion about constant corporate surveillance. Will Amazon be able to resist saying, “I see you’re flushing a lot…do you need Lomotil?” It’s also likely that this effort will end up with a lot of partial standards, with many incompatible products bearing the logo, and standards that don’t really standardize anything. But it’s worth remembering: the internet could never have happened without open standards. Neither will smart homes.
- Jack Dorsey is proposing the development of an open standard for social networking, lead by Twitter (under the handle @bluesky). Reactions have been very mixed; Techdirt has the best analysis I’ve seen. Tim Bray’s appropriately skeptical take is also very good. Certainly, we’d be in better shape now, though, if Twitter hadn’t decided a few years back that it didn’t need its developer ecosystem.
- Fred Benenson mentions topic-focused micro social networks in a tweet. This is a theme that I’ve seen coming from several directions lately; I’ve seen Mastodon explained as a collective of topic-focused networks, it’s a possible direction that Twitter itself could take with Dorsey’s proposal for open social network protocols. If Facebook’s hegemony is going to end, this may be how. The only way to compete with a monopoly is to change the game.
- There are also proposals for open standards for machine learning operations and governance coming from Cloudera. It must be open standards month.
- It’s always a safe bet that cybersecurity is going to get weirder. There’s a lot of angst about deep fakes and phishing, but the biggest struggle will be with attacks against AI. Nevertheless, the ground truth with computer crime is that “criminal” isn’t a job you get by being smart. The biggest problem will continue to be the annoying, primitive, and lucrative attacks that companies haven’t responded to for the past decade.
- Here’s a commercial implementation of a technique I suggested for more secure biometric authentication: the “burner fingerprint.” My suggestion had a more subversive flavor (3D print a “fingerprint”; use that for ID; when concerned about device safety, such as when traveling through customs, throw the fingerprint in the nearest trash can; print a new copy at your leisure). This doesn’t quite go all the way, but they’ve clearly got the idea.
- With Braket, Amazon adds itself to the short list of vendors that are making quantum computing available in the cloud. Amazon doesn’t have (and isn’t starting) its own cloud effort; they’ve bought machines from existing vendors that represent different approaches to quantum computing (D-Wave, Rigetti, and IonQ). Whether or not these are the right approaches, Amazon’s thesis that most quantum computing will be in the cloud is almost certainly correct.