comments on "these guys"

[Dave Farber <dave () farber net>]

Begin forwarded message:

> From: Scott Alexander <salex () dsalex org>
> Date: October 18, 2009 11:49:10 EDT
> To: "David P. Reed" <dpreed () reed com>
> Cc: David Farber <dave () farber net>
> Subject: Re: comments on "these guys"

> I agree in part and disagree in part.  Firstly, I don't understand  
> the constraints the PMs are currently working under.  To some  
> extent, I'm not sure if they do either.  My impression is that an  
> attempt is being made to fix what's broken at DARPA, only there  
> isn't consensus on what parts were broken.  Of course, it's been  
> publicized that Regina Duggan is touring universities attempting to  
> re-engage them and to recruit new PMs.  And, of course, it's really  
> her and her bosses'  perception of what's broken that will be the  
> goal of the fix.
>
> So, I don't understand enough to say anything about DARPA now.
>
> OTOH, I think the PMs had a fairly wide autonomy to pick the  
> problems of interest to them.  Those problems then had to be framed  
> in a way that spoke to Tony Tether at the new start.  In particular,  
> the military relevance had to be clear and it had to be clear how  
> success would be determined.  I think this was, in part, a reaction  
> to the perception by some that in the 90s, it was possible to get  
> money for just about anything (if you knew the DARPA handshake) and  
> that there was no accountability.  But this isn't a new problem.  If  
> you see http://special.lib.umn.edu/cbi/oh/pdf.phtml?id=146 , George  
> Heilmeier describes having seen a similar problem in the 70s and  
> having taken over DARPA with some of the same goals that it sounds  
> like Tether had.  I've also heard that there was a significant focus  
> on making sure that DARPA was bringing value to DoD at a time when  
> soldiers were being killed.  The phrase I heard was "will spending  
> this money be better than spending the same money on body armor?"   
> Within those constraints, what I've heard is that there was pretty  
> wide latitude.
>
> As originally framed, IAMANET was asking about how to fundamentally  
> redesign the network with security as a primary goal.  Chris Ramming  
> made comments about how it had taken 40 years for IP to come into  
> being and get to where it is today, so what should be the course of  
> the next 40 years.  My impression is that those sorts of programs  
> were harder to sell than at other times, but certainly not impossible.
>
> (I think "performers" comes more of the defense contractor  
> perspective than how you might be interpreting it.  I've always  
> heard it used as one who performs (or executes) on a contract.  Of  
> course, that goes along with the notion that DARPA is about  
> carefully planning innovation.)
>
> In terms of advising PMs, you're quite right that the BAA is too  
> late.  The BAA is a legal solicitation and to make substantial  
> changes would require canceling the BAA and releasing a new BAA.   
> The PMs do have some ability to interpret a program within the  
> constraints of the BAA, but straying from the BAA is difficult.   
> Trying to offer advice at that point is like showing up at a house  
> closing and trying to negotiate the terms.  Of course, submitting a  
> proposal also allows some ability to sway things.  The contract,  
> particularly the SOW, replaces the BAA.  Thus, it is an opportunity  
> to convince the PM of something about the approach to the program,  
> albeit a blunt instrument applied rather late.
>
> Seeing a PM before a BAA comes out implies that you need to either  
> be the first to present a good idea to him or that you need to hear  
> about something that will be coming out.  The latter can be a bit  
> incestuous because PMs, like everyone else, tend to go to the people  
> they know for advice.  That said, PMs do have a fair ability to cold  
> call people.  Apparently if you give out money in units of $1M,  
> people return your calls quite quickly.
>
> It's also worth keeping an eye on workshops.  I was involved in the  
> Assurable Global Network workshop.  I believe that that workshop  
> significantly affected Chris Ramming's thinking for IAMANET.  I've  
> also seen a number of PMs who will talk about upcoming programs  
> given the slightest opportunity.  Of course, in part, their goal is  
> to make sure that they have a wide variety of people who will  
> eventually bid on the program.  In part, though, that's a way of  
> broadly soliciting ideas.
>
> I'm also told that at the end of the first year, a PM has so many  
> good ideas that they don't need any new ones for the rest of their  
> term.  Thus, if you present them with a good idea after that, it has  
> to displace another good idea.  Speaking to new PMs does present one  
> with more opportunity to influence what DARPA does.
>
> I've also found that the non-miltary PMs tend to have very good  
> career opportunities after DARPA.  Of course, any of the defense  
> contractors are interested, just for the insight into how the DARPA  
> process works and for the contacts.  And, as you and I have both  
> observed, the PMs tend to be very smart people and so are also  
> recruited on that front.  (I exclude the military PMs above only  
> because I don't understand military careers and so can't speak to  
> those effects.)  Beyond that, Tony Tether had said that he figured  
> the point of having a term appointment was so that no one would be  
> concerned about making a career at DARPA.
>
> Personally, I think the biggest challenge for DARPA is finding a  
> middle ground.  How do you build a system where there is  
> accountability for the money that is spent that encourages the level  
> of risk needed for great breakthroughs?  I wouldn't want to see my  
> tax dollars go to those who know the DARPA handshake even if they no  
> longer produce anything.  I also believe that we've recently seen  
> cases where meeting the metrics became the goal rather than solving  
> the problem.  And, of course, the problem has to be solved in the  
> context of the government contracting system.
>
> This is a bit more long-winded than I'd intended.  And I do think  
> that there will always be things broken in an organization the size  
> of DARPA (particularly while its part of an organization with the  
> size of the DoD).    I've just found that my experience doesn't  
> match what you describe.
>
> Best,
> Scott
>
> On Oct 18, 2009, at 8:03 AM, David P. Reed wrote:
>
> > Scott - I did not mean that DARPA in general or the specific PMs  
> > are idiots - far from it!  When I said "these guys", I was  
> > referring to the system they operate in.  In the area of making the
> >
> > By advising, I meant something very specific: advising on the  
> > *framing* of problems by the DoD by people who have a broad  
> > perspective and are not entangled in the perverse incentives of the  
> > DoD community.  There's not a lot of advising of this sort that is  
> > possible.
> >
> > I've been part of the "advising" process as a member of ISAT, so I  
> > know that there are some channels for advice. But almost all of the  
> > advice actually given in such contexts is narrow, technical, and  
> > unable to challenge the cultural assumptions and framing of the  
> > larger issues that the military faces.  For the last decade or so  
> > this was really problematic: the management style was to shut down  
> > any creative exploration of the problems, in favor of setting very  
> > tight constraints, deadlines, and competitions (why are the  
> > contractees called "performers" unless they are viewed as puppets  
> > who have very little scope of action beyond performing to the tune  
> > of the PM).
> >
> > For example, BAAs and proposal days are fora where offering advice  
> > of that sort is a waste of time, and likely seen as rude.  And once  
> > a contract is awarded, redefining the goals to solve the "real  
> > problem" is fraught with the problem of "missing milestones".
> >
> > DARPA PMs these days have very limited scope of action. They have  
> > to worry about their careers, whether they are academic, military  
> > contractor, or military in the long run.  Careerism is a major  
> > problem in the DoD technology space as a whole, and it's become a  
> > major problem in academia over the last 20 years or so.  Would I  
> > advise a PM to question the wisdom of the way a particular military  
> > problem is framed by a branch of the military?  Would I advise a PM  
> > to invent the Internet of today and seek to implement it?  It would  
> > be a huge risk to their careers, whether they are academic or  
> > military.
> >
> > Thus, "you can't advise the guys".
> >
> > - David



-------------------------------------------
Archives: https://www.listbox.com/member/archive/247/=now
RSS Feed: https://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com