Monthly Archives: March 2014

I think I can be forgiven if, even with very limited data, I make the observation that coincident with a new CEO much is being revealed by GM. Logicing (not a word but I like it; sort of like spelunking) our way through, it could be the case that this stuff just happens to be coming out. Alternatively, it could be coming out either because of forces within GM’s control or outside that control. Given that by GM’s own indications the ignition problem was known since 2004 (possibly 2001), it seems improbable that some force outside its control caused these revelations. After all, the fatalities have been known about since the time of each occurrence. Taking it one branch further, if this is under GM’s control, we can then assume active involvement by GM in doing things that brought this all to general attention.

All that by the way is something of a side story, but one where we learn something about leadership by the new CEO for GM. (I think the degree that you believe she has put this on the table and dealt with it would correspond to your notion of her positive leadership) The main story of interest is where the previous management team is and how they will be impacted by this. Certainly the current CEO was somehow in that structure and it is an interesting question to know what she knew and what she could do about it. In a simple sense though my question is who, if anyone, will see claw backs or other sanctions as a result of this? Having a major recall due to a defect that happened on your watch strikes me as something that should result in negative outcomes. My initial thinking was that a corporation’s only recourse against management would be financial (clawbacks) but a few minutes of pondering lead me to think that firing is also a recourse that should be on the table for those who are still at the firm. Recourse strikes me as continuing up through and including the board. If the board is saying they didn’t know about it you assume inadequate oversight by the board. After all, defect causing major recall seems like the kind of thing the board should know about whether management is telling you or not. If they did know, then they strike me as at least as culpable as anyone else in the chain.

This gets a bit into theory of the firm where in this case the firm has been damaged by the actions of its management. If we simplify and assume a single owner it is much easier to come to the conclusion that this damage would be rectified (at least by a rational owner) at this late date by recourse against those responsible. You would assume that such an owner would have known sooner and acted sooner unless there was a compelling incentive not to have acted (always possible). With fragmented ownership and a board there is that same issue of compelling interest not to have acted OR the issue of not having sufficient interest in the long run of the firm. A theoretical single owner who owns today and will own 20 years from now only has to decide in which time period to take the hit (leaving aside obvious moral obligations). For a board though, deferring to a later period may be a viable and rational strategy even if it makes more sense to take the hit in the current time frame for the firm. (again leaving aside moral obligations).

I don’t know what the solution is for the mixed incentives and at least partial alignment of interests with management and the board as opposed to what can be thought of as the firm’s interests as an entity.

PS I think there is a fair criticism of corporate management systemically when nothing comes from (and in some sense can come from) circumstances like this.


I’ve been thinking about an earlier post on experts and the method described to try to overcome communication\clarity issues at the expert level. Rather than rehash, please see that post.

Would a method like the one proposed work in a corporate setting? In an academic setting, it seems doable but somehow unlikely. In the business world it goes from unlikely to improbable for several reasons. Let’s try to do a one to one transfer to a business setting. First, who plays the role of students? It is one thing to act as a student in class but what employee really is going to want to have to step up and show they don’t know something, possibly in front of their boss? Second, who plays the expert roles? You probably have someone who could do the instructional expert role. In the case where some important decision is being made there should be someone who is championing it and has a good grasp (though this is by no means guaranteed). What about the other experts who are supposed to be listening in as experts to gain insight? What is their basis for being an expert? In some cases they may be qualified but in many they won’t. There is an assumption that executives should have developed expertise across a number of domains but that certainly isn’t factual. This is both an internal and external assumption in that both the executive and those around him/her expect it to be the case. Even for those who have some grounding, is it sufficient to be called an expert? This reinforces the need for going through the basics work prior to getting to the important point while also reinforcing that going through the basics is going to be viewed as insulting. Third, how do you get the time to make a method like this work? Even in a single area in a firm it can be difficult to get time to go over the whatever the important point is without the basics. How do you get enough executive time to add to that to go through the basics? Fourth, how do you know when the point is sufficiently important to go to this level of effort?

The problem space for the general problem of lack of clarity due to insufficient basics work caused by the potentially insulting nature includes ideas like: How do you get people to let go of perceived expertise and allow them to work through the basics? and How do you get them to participate at all?

I don’t have a good idea yet for a workable solution in the business space. Like the method above, the ideas I come up with seem unlikely to prove workable.

PS – This isn’t to suggest that there are no cases in business where the basics aren’t reviewed prior to getting to the important point.

PPS – I do recognize that it is rational for an actor to want to protect status in a given situation. That is part of what I am referring to in the problem space of ‘allowing’ working through the basics.

PPPS – If this is all a bit mucky consider a simplistic example in which a CFO is making a case for altering the financial structure of the firm. The important point is whether new structure X is better than existing structure Y. Is it likely in this case that the C level staff (and maybe the board) would do anything other than debate that particular point? Now consider the academic method discussed in the earlier post. Who would be willing to act as the students (employees?) in this case as the C level staff sat on the sidelines (interjecting as experts) while the CFO took the students through the basics of financial engineering specific to this decision prior to explaining his/her rationale for why structure X is better than structure Y. It just doesn’t sound very likely. What alternative method would achieve the results?

I’ll preface this by saying this is anecdotal and speculative.

We were talking about diet and exercise at a dinner with friends the other night and one of the points of discussion was an emphasis on diet in terms of weight loss. A point that came up that attempting to lose weight isn’t recommended at the same time as training for endurance athletic events. (Long distance swimming, tri’s, marathon/ultra, etc) The thinking is that attempting to lose weight during the training cycle will likely reduce performance both in training and at the event while being difficult to have any success at due to increased appetite and decreased willpower. This fits in with some noodling I have been doing on multiple equilibrium points for weight loss.

Multiple equilibria refers to a situation where a system has more than one point where it can rest. I’m bastardizing that concept, perhaps poorly, in this case to indicate multiple points of optimization of an outcome. In this case, multiple different ways for weight loss to occur at a ‘better rate’ though all might not be equal. This isn’t intended to be a complete list. The reason I think the notion is important is that multiple equilibria might explain part of the contending (and confusing) possibilities for weight loss.

Anecdotally, I can think of several situations where weight loss seems to have been more optimized in my case. Long duration\moderate load\high frequency\high calorie – this is a case experienced in thru hiking. Thru hiking is a long distance hiking excursion lasting weeks or months. Daily hiking occurs in the 6-10 hour range. Load factors are pace, backpack weight, and altitude gain/loss. Caloric intake is high. In this circumstance, there almost isn’t a caloric intake reconcilable with holding weight. Sufficient caloric intake often becomes a problem. Depending on the load factors (it is possible to go slow enough, take enough days off, lower hours hiked etc) to maintain or increase weight. Many people lose weight in this combination of factors. Light to medium duration\light load\high frequency\low caloric intake – this case is experienced where diet was a focus and exercise was limited to walks lasting about an hour 5+ days/week under a light load. Load factors were normal pace, one hill, work back pack. The key here was that the level of activity was light enough to maintain a lower calorie diet without failure. Frequency of activity was high.

Contra example: Mixed duration\mixed load\high frequency\lower calorie – this is the case in marathon and ultramarathon training while trying to get lighter. Duration tend to be an hour plus with long day duration 3 – 6 hours. Load factors are types of training (track workouts, tempo), pace, elevation gain. Frequency is typically 4-6 days a week. Caloric intake is a balancing act between sufficient energy for workout performance and muscle maintenance while trying to create caloric deficit sufficient for weight loss. This did not lead to weight loss. The line to balance is too fine. Individual workouts produce different caloric deficits that a generalized daily caloric load doesn’t match. Mental fatigue and mismatch lead to failures.

So where is this going? My current thinking is that weight loss for an endurance athlete is probably best achieved outside a training cycle where diet is the leading focus with exercise dependent on maintaining the diet. Once a set level is achieved, then moving onto a training cycle with a goal of weight maintenance. For the non athlete, I suspect that the best outcome is dietary focused with frequent light to medium duration and load exercise again dependent on diet maintenance.

In both cases it would imply getting comfortable with a dietary pattern that can be maintained and leads to weight loss and only then bringing exercise levels up to but backing away from any level that challenges dietary maintenance (staying on the diet) or causes failures. There is some non zero level of exercise that should always be reconcilable with a maintainable diet (walking slowly on a flat surface) but the emphasis would be on care in dialing back should conflicts arise.

This is all purely speculative of course.

It is always interesting when you experience a case of co-development. Sort of like a case of deja vu. Co-development being the circumstance where more than one person develops roughly the same idea independently. It isn’t necessarily of any importance who develops it first; it could be across a great space of time. The critical factor is whether there is true independence of development. By that I mean that the idea is spawned without any cross contamination of the process by information from other instances of it having arisen.

Recently I was working with a friend who was developing course content and he was lamenting the fact that the group who would be participating in the course might not be those most in need of the course. He knows factually that the participants will be more junior personnel, who will of course benefit, while simultaneously being aware that the senior personnel at their firms likely have limited exposure to the area. They are more likely to gain from the course given their deeper experience and exposure and more urgent need for the material. The conflict is twofold. The senior folks can’t be seen to not know the basics either by their juniors or by their peers. So it isn’t as simple as having a separate class. The possibility of loss of face will keep them from participating.

As we talked about the problem, a notion occurred to me about having the senior staff attend on an audit basis. They would ostensibly participate to observe how their staff were doing without being subject to the assignments and participation that might reveal their shortcomings while still being able to hear the material and participate to the degree that works for them.

How does this relate to co-development? In the last week I have been reading Daniel Dennett’s book Intuition pumps and other tools for thinking. He has a chapter on using laypeople as a communication device among experts. He notes that among experts, the tendency is to under explain when speaking to a fellow expert. The notion is that it is insulting to over explain (go through all the detail needed to build the foundation for a particular point). Solutions that ask experts to be overly explicit fail in Dennett’s experience. His idea is having experts sit on the sidelines while students are taken through the basics leading up to and including the point. This allows the experts to go through all of the detail (and clarification of understanding) without being insulted by the process.
It is entirely possible that my notion was somehow influenced by something that I came into contact with either from his development of the idea or that of someone else. Nonetheless it is also possible that we developed the idea independently.

PS This seems to point out a clear cut case of a persistent mental error. Being unwilling to go back through the basics because it may be revealing of a weakness strikes me in no other way. It may be a rational actor response to one in that situation in some sense.

Scott Adams recently had a post where he stated that common sense was a function of experience. While I agree that experience is a component of common sense (CS), I think that it is a lot more complicated. Anecdotally, there are plenty of people with lots of experience that seem to lack CS.

So what is CS? That is a pretty tough question. Google gives results that stress forms of judgement and practicality in common matters. That strikes me as covering a lot of ground. I’m thinking it is both a relative and malleable concept. One person’s CS might not be another’s.

There are a lot of pieces I’d throw on the table of which, experience is one. I did X and got Y result in the past. Now it is a question of pattern recognition -> is X occurring? To that, I’d add other people’s experience. I trust person Q, they did X and got Y result. That dramatically expands the number of CS scenarios to which you have access. Again, pattern recognition. But is it all backward looking? I’d argue no. There are additional pieces. Here are some, though this is by no means a complete list. Forward thinking or alternatives analysis, both of which I’d put under the heading of imagination, is a piece. I can do X. What possible outcomes of X are there and how likely are they to be the result? If none of the outcomes or none of the probable outcomes are results that I want, then don’t proceed. Forward planning. I’m going to do X, and the likely results are Y, but if not Y how will I respond to alternatives results? Skepticism. I have good reason to doubt X or little experience with X. I should proceed with a much higher level of caution in regards to X and taking more opportunities to consider how to proceed. Patience. Is there a compelling reason to proceed with X right now? If not, why not delay? More Information may change the result. Observation. I see these details about X which informs me about it. How does that new information play in to my consideration of X? Attention. Does this item have my attention or for some reason (tired, hungry, etc) am I not able to fully think about it?

This all sounds rather complicated what with using variables in the examples, but there are quick and quite simple example versions of all of these that I think sound like instances of practical judgement(CS). I think it is just a question of scale.

Which leads me to where I think I am headed with all this. Common sense is evidenced in the use of a broader set of thinking tools/skills, though perhaps only in their simplest incarnations.

This topic deserves a more extensive treatment but this is what I have for now. More someday.

It’s right there in the name, password. It is the long running solution for the problem space of trying to identify a specific individual to allow other actions to take place. For all the talk of eliminating passwords and two part authentication, we still haven’t done a very good job of designing for the human element. Humans are notoriously bad at remembering certain types of information. We do better when we can chunk together information. The escalation to more complex types of passwords due to better and better models for breaking them tends to break down our ability to chunk. (Did I use a capital H or lowercase?, what was my special symbol?) We tend to repeat passwords, use ones that are too simple, or write them down, all of which hammer home that this is a design problem in human memory.

So why don’t we move to something that we are good at remembering? Our visual memories tend to be excellent, with a large part of the brain devoted to this function. People that compete in memory contests leverage this capability to build associative visual pathways to then attach other items they are trying to remember. Why not leverage this as a means to identify us? I’m thinking that as part of the process instead of creating a password, I load some number of pictures to the site to be used for identification purposes. When I return to the site, one of my pictures is presented with say 14 other pictures and I have to pick the right one. Maybe I have to do this more than once. I’m guessing my accuracy at picking would be nearly perfect. The pictures can be continually randomized as to placement on the screen so a click in one area doesn’t necessarily follow the next time through. I can think of a ton of other ways to enhance/alter the process to increase complexity without really changing the human ability to identify the correct one. Moreover, even on a site that isn’t used for years, my odds are still pretty good at remembering things in the picture. There is no reason this shouldn’t work on a phone or a laptop as easily as on a website.

If you chose to implement, please just send me my 20%.