s3e05: Not A Tightrope

by danhon

0.0 Station Ident

2:19pm, sat in a grocery store cafe again with a Diet Coke and some really tasty sea salt caramel chocolate things. Queen’s I Want It All is playing on the store PA. It’s super sunny outside, no clouds at all to be seen and there’s a butterfly playing chicken in the road.

1.0 Not A Tightrope

There is a thing about me where success or achievement isn’t something that can be celebrated or acknowledged in its own right because success just means… you have to succeed again. Getting past a feasibility study means there’s more pressure on the first stage of implementation. Getting past the first stage of implementation means there’s even more pressure on the second stage, and so on. Success at things isn’t a re-set with each opportunity, it’s a succession of walking on a set of finely calibrated tightrope walks, or ascending to the summit of everest. It doesn’t matter how high you’ve gotten or how far you’ve walked because the next bit – and there’s always a next bit – is always more important and invalidates everything else that you’ve done so far. It’s not like you get to walk on a tightrope or do part of an ascent and save your progress at a basecamp. It’s that each move forward is an opportunity for failure. Which, let’s be frank, isn’t a particularly compassionate way to deal with your self.

– chalk another one up for the “integrating systems is harder than you think it’s going to be” camp: a new report by NASA’s Inspector General[1] via Ars Technica[2] finds that the decade-long development of their Spaceport Command and Control System software is 77% over budget, over a year late, and with at least 2.6m lines of code being called “glue-ware”. Money quote is “it is now clear [NASA managers] underestimated the complexity of the software integration activities that would be required.” There exists, apparently, commercial off-the-shelf launch software, used by Orbital Sciences and SpaceX. Part of the problem looks like in the *ten years* since NASA made the decision to go down this path, there wasn’t an annual review to see if anyone had made anything better. Which, apparently, they have.

– internet-acquaitances Sara Wachter-Boettcher and Eric Meyer have written an in-my-humble-opinion must-read book on design and empathy called Design for Real Life[3] which is probably better than the book I was trying to write about design and empathy because it has actual advice as to how to combat such an empathy gap. Wachter-Boettcher has done an interview[3] with A List Apart.

– now that the first few New Wave of virtual reality headsets are here complete with stupendously powerful consumer graphics processing hardware, you might be wondering: how can I get something that will fit all of that in one of those cool-looking pelican cases that I can travel with? Leonard Lin has got you covered with a build list[5] that includes weights of components. On the whole VR is A Potential Next Big Thing, I’ve been talking with a number of friends about what the equivalent now is of a parent borrowing a computer from work in the mid 80s/90s. It’s not necessarily a hardware thing, although 3D printers might be a candidate, as well as VR, it could just be a software thing “the web” and “making and selling your own software” and it may just well be that bringing home and exposing to young children a bit of kit *at the same time as a multi-decade economic transformation* just might not happen again. At the same time, it didn’t stop me from getting the 3 year old/me one of the Sphero BB-8 units to “teach him programming sometime” now that you can use Scratch with them.

– my work with helping to “fix” government now involves bidding on government contracts which means I’m now on the receiving end of requests to do things like “find your university degree certificate and scan it in” and “get written references that you can do the sort of things that we want you to do” as well as “please provide a 5-10 page description as to how you’re going to do the thing that you’re going to do”. Fortunately I’m much less good at procrastination these days.

– I’m seeing a few weak signals again (precipitated, I think, by the renewed interest in AI post AlphaGo) about What We Can Do When Artificial Intelligence Appears, some of which point to existing writing about Artificial Intelligence Is Already Here And The Equivalent To The Paperclip Maximising Function[6] Is Maximising Shareholder Value[7, 8, 9].

[1] (trigger warning: PDF) https://oig.nasa.gov/audits/reports/FY16/IG-16-015.pdf
[2] NASA launch system software upgrade now 77% over budget | Ars Technica
[3] A Book Apart, Design for Real Life
[4] Design for Real Life – An interview with Sara Wachter-Boettcher · An A List Apart Article
[5] The Most Portable VR Workstation – random($foo)
[6] Paperclip maximizer – Lesswrongwiki
[7] The Singularity Already Happened; We Got Corporations | Quiet Babylon
[8] Omniorthogonal: Hostile AI: You’re soaking in it!
[9] Invaders from Mars – Charlie’s Diary

2:40pm, and it’s time to go meet someone for coffee. Not a great day today, but one that will end nonetheless, and there’ll be a chance for a do-over tomorrow.

Best,

Dan