The h Bar

vzn

00:17

dmc re "mismanagement". the New Yorker article re fusion makes it clear that even recent large projects have massive management/ political challenges, its not unique to any one project. modern LHC has succeeded but at significant cost. imagine/ suspect there is significant politics, hurdles, cost overruns there also but so far its fairly well hidden.

humans have difficulty doing large projects, its not so much the technology but the collaboration/ political structure that is extremely challenging.

it seems nobody has invented a highly effective political structure so far. neither democracy/ authoritarianism work all that well. and there do not seem to be any other models.

eg there are decades of failures of large IT projects for example, they are not well recorded in history, although there are a few rare ppl studying/ recording them. one old book on the subject was called "death marches".

alarge

00:38

@vzn I think there's a lot of literature, especially on IT project management. For example, the classic The Mythical Man-Month (and of course modern literature is filled with techniques such as full-stack, devops, agile, lean, scrum etc. which all show the shortcomings of the alternatives, usually with several examples). And then there are websites like the daily WTF.

dmckee

After the SSC debacle, the DOE established a more detailed set of goals to be met before getting access to money. CD-0 through CD-5 (a few of which have sub-goals).

There are also additional accounting requirements and training that the senior money managers have to take.

vzn

there may be some "lessons learned" but the real issue at the core is large scale human collaboration and it will possibly never go all that smoothly... aka bureacracy

dmckee

Various half-a-billion dollar projects have completed with no worse overruns than the rest of the federal government (which means medium bad ...).

user54412

@vzn My understanding is something like 70-80% of the funding for that is coming from NASA. In a way it's no different from the Apollo era -- much of the hands-on manufacturing is contracted out to private industry. It's just that these hip new startups get all the public attention today, whereas NASA got all the credit back in the 60s.

vzn

wrt "big physics" & spaceflight etc a half-billion dollars is a small prj so to speak.

yes have read some of the IT prj mgt literature. its funky that there is a strong trend toward less mgt there eg wrt "agile" etc.

dmckee

00:41

Certainly there is simply friction built into big projects, but there is bad (which is what we have now) and simply incompetent bordering on negligent which is what went on during the SSC.

I'm in particle physics. 1/2 a billion is big enough to bring on the most stringent control measures.

vzn

agreed SSC may have been worse than average, its possible... but can the US point to any successful prj on that scale? we can barely pull off "obamacare"....

alarge

@vzn Well it's not all just bureaucracy. There was a discussion earlier this week about the importance of social skills in a technical environment, and in a sense this connects to that.

vzn

cost control is not the big solution to large prjs, its certainly a big part of it....

user54412

@vzn And I don't know much about the LHC, but I can tell you that in astronomy (Gaia, Euclid, etc.) the ESA limits how much the US can contribute -- they don't want the US to individually contribute more money than all of the actual European countries individually.

vzn

look at how many billions of dollars were squandered in iraq in the last decade.... its staggering but still fairly well hidden....

(rats meant to edit that & deleted it lol, tiny-close buttons on my iPad mini lol)

its deeper than mere social skills, the core problem involves large-scale mgt/ control/ organization/ accountability/ political structures. even large govts routlinely fumble large prjs.

limiting individual nation contributions may help some... but then theres the big challenge of intl cooperation. LHC possibly is one of the most successful models ever.... maybe they should document their org structure/ system somehow....

obamacare opening day all the web sites crashed... and for weeks.... really wanna say something nice but not very impressive :(

user54412

00:52

@KyleKanos There are several bets in my office on whether or not JWST will ever return data. I always take the position that it won't.

2

vzn

lets look at large corps as organizational systems. sometimes dysfunctional. MS recently announced layoffs around 15%. IBM (something like 400K employees) ... strong rumors of a staggering 25% layoffs upcoming.

user54412

It's not clear how much JWST failing will impact NASA as a whole, since it's actually just a small fraction of NASA's budget, but I'm pretty sure failing to deploy will make sure no money is spent on astronomy proper for a number of years. If it fails, the best thing for astronomy is for it to blow up on the launch pad, so we can blame the rocket scientists :P

alarge

@vzn Well there's been huge shifts in their industries and business models, so it's not surprising that layoffs happen.

vzn

the shifts have happened for years and the organizational brains of the systems failed to adapt.

because they are like dinosaurs, large bodies but small brains.

and nobody has really figured out how to have a big body & big brain. although maybe LHC comes closer than ever. hubble telescope is another "general" success although look how they botched the lens, holy cow... & saved it with a legendary Hail Mary spacewalk.... most complex ever....

but hey we do sports pretty well eh? Super Bowl? :)

user54412

Someone connected to Hubble recently described it as the best story you could come up with for a movie. All the difficulties and doubts getting it up, the moment when everything seems lost, the plan that relies on a miracle, and the final success, creating some of the most stunning images in human history. It just sounds like a screenplay.

vzn