I agree that we won’t get hyperspecialization, but I don’t think hypergeneralization is going to happen either. Deliberate practice/10,000 hours still matter to adding value.
What I personally think will happen is de-canonization. Specialist/generalist are relative terms in some sort of collective social ontology of professions. Something that allows you to collect statistics and segment the labor market. What’s happening is that more people are becoming illegible and undefinable in traditional terms because their skills and knowledge, whether generalized or specialized, simply don’t fit any normal categories.
Rita Kop (Welsh Cloggy)says:
You are right.The fragmentation of knowledge by high levels of specialization is a step in the wrong direction. Midgley highlighted: ‘Traditionally, the value of knowledge centred on understanding – on the power to see the connections of things, to wonder at them, and so to live wisely’ (Midgley, 1989, p. 3).This indicates the need not only to investigate deeper one particular issue, but also for interdisciplinarity and the making of connections to not only understand this particular phenomenon, but also its relation to the world.
Alejandro Weinsteinsays:
Seems XKCD has an appropriate comic for everything:
I think that everything is linked: thanks to specialization, we went further on specific topics, and thanks to books, then the web, we developed knowledge in a way than more or less anybody can access… Thus this ability to go toward what you call “hypergeneralization”.
I rather believe that humanity is going forward on the two directions, one supporting the other. But this generalization phenomenon is indeed the most impressive and the most evolving one, turning everyone into a potential expert on almost any topic.
I’m not sure of the consequences of all this, but I would say that if the essay’s idea is quite right when it comes to research (unless generalist geniuses like Einstein discover new specialties!), we indeed tend to be generalists in our everyday life.
Engineers at Facebook or Google already have much autonomy. They often get to run many of their projects. If you are going to build GMail, you better have a broad expertise. The autonomy to build significant things requires that you to be a generalist at times. Hyperspecialists must be constantly managed. Those are not the employees Google and Facebook are looking for.
those who perÂsonÂalÂly genÂerÂate inÂtelÂlecÂtuÂal capÂiÂtal [must be] treatÂed as manÂagers in their own right, and posÂsess the exÂecÂuÂtive auÂthorÂiÂty reÂquired to deÂcide how best to deÂploy their time.
I think that the future belongs to those who will be able to quickly master new things. The so-called specialists will face mounting pressure as new non-specialists start to occupy their territory with greater and greater ease.
Itmansays:
Unless you manage, you do must have a certain specialization. Managing is a different issue, the technical knowledge does not matter much. What matters is the ability to bullshit confidently.
Itmansays:
Daniel,
I think we define generalization too differently. Actually, I was working in a small team that was delivering a web-based e-mail to millions of people so that I could see what people were doing. Even those who were doing a lot of different tasks, in my opinion, specialized in a very narrow field. That is they had experience with C++, Linux, SQL, perl, mail protocols, and some search algorithms. If they were generalists they would also write phone aps, design IDE for Windows, MAC, etc…
Your future wealth is determined by how much you can expand your mind beyond the capacity of your biological brain, not by your current skills.
There is nothing risky about this proposition.
Itmansays:
It is definitely easier. It is just not feasible unless you manage. I am afraid of the perspective when everybody is becoming a super-generalist (without deep technical expertise) and who is doing the business? I want a plane to fly smoothly instead of crashing. Just because somebody was catching up with the experts too quickly.
Itmansays:
Daniel, I completely agree with you on the statement. Your or mine wealth does depend on it. However, this model is not scalable. If everybody starts to generalize, the civilization will collapse. Well, and anyways if I, e.g., don’t like overgeneralization, I would prefer to stick to my technical skills and be happier despite it would mean less money.
(…) questions asked required specific knowledge such as what TCP port a service ran on or what tcpdump command was used to print certain types of packets.
Ironically, you can google these questions very easily.
These are not questions I would ask a potential new Google engineer.
Hmmm. I didn’t see much evidence of generalization in the Marcin
Wichary article. The subject of the article is primarily a user
experience (UX) engineer who leverages various web technologies
(e.g. HTML5). I didn’t see much evidence of cross-disciplinary
involvement, even within software engineering (e.g. operating systems
or AI).
My experience tends to match Itman’s, especially with regard to
Google. The interviews I’ve had with them have been very specific and
detailed; questions asked required specific knowledge such as what TCP
port a service ran on or what tcpdump command was used to print
certain types of packets. Not the higher-level conceptual engineering
types of questions I was expecting. On a more general note, most
people are not capable of being outstanding in several subfields of
the computer industry, so they are forced to specialize in order to
pass an interview such as the Google interviews I’ve had. Managers
benefit from generalization more, because it enables them to employ
multiple strategies to run their teams, but individual contributors seem
to be judged on highly specific topics such as how to code a particular
algorithm, etc.
Perhaps Google has changed their interview strategies, but I haven’t
seen much evidence of it from other sources of information on Google
interviews.
Itmansays:
Daniel, it probably depends on a department. The last time I talked to a Google recruiter, ironically, they said they wanted a generalist. What they probably meant, however, was that the person would know C++, Unix, network protocols, databases, and a bit of the WEB (as opposed to, e.g., SQL coder), rather than being a computer scientist, a successful coder proficient in Linux/Windows/Mac technologies, who is running his own business in spare time.
JackBsays:
Even when technology enables easier access to knowledge doesn’t mean there wont be hyperspecialization. Even if we are expanding our minds exponentially, the knowledge base in the world is expanding even more exponentially! Because of this increasing distance between the general knowledge base and our brains ability to handle information, you have to focus on a narrower area in order to do any meaningful contribution to this knowledge base.
Not saying that generalists doesn’t have their place too, but more in the role of creating applications with existing knowledge. So I wouldn’t say it’s more risky to work as a generalist than specialist.
@Alejandro
Indeed.
I agree that we won’t get hyperspecialization, but I don’t think hypergeneralization is going to happen either. Deliberate practice/10,000 hours still matter to adding value.
What I personally think will happen is de-canonization. Specialist/generalist are relative terms in some sort of collective social ontology of professions. Something that allows you to collect statistics and segment the labor market. What’s happening is that more people are becoming illegible and undefinable in traditional terms because their skills and knowledge, whether generalized or specialized, simply don’t fit any normal categories.
You are right.The fragmentation of knowledge by high levels of specialization is a step in the wrong direction. Midgley highlighted: ‘Traditionally, the value of knowledge centred on understanding – on the power to see the connections of things, to wonder at them, and so to live wisely’ (Midgley, 1989, p. 3).This indicates the need not only to investigate deeper one particular issue, but also for interdisciplinarity and the making of connections to not only understand this particular phenomenon, but also its relation to the world.
Seems XKCD has an appropriate comic for everything:
http://xkcd.com/903/
I think that everything is linked: thanks to specialization, we went further on specific topics, and thanks to books, then the web, we developed knowledge in a way than more or less anybody can access… Thus this ability to go toward what you call “hypergeneralization”.
I rather believe that humanity is going forward on the two directions, one supporting the other. But this generalization phenomenon is indeed the most impressive and the most evolving one, turning everyone into a potential expert on almost any topic.
I’m not sure of the consequences of all this, but I would say that if the essay’s idea is quite right when it comes to research (unless generalist geniuses like Einstein discover new specialties!), we indeed tend to be generalists in our everyday life.
@Itman
Engineers at Facebook or Google already have much autonomy. They often get to run many of their projects. If you are going to build GMail, you better have a broad expertise. The autonomy to build significant things requires that you to be a generalist at times. Hyperspecialists must be constantly managed. Those are not the employees Google and Facebook are looking for.
This is broadly applicable. I’m with Dorian Taylor:
those who perÂsonÂalÂly genÂerÂate inÂtelÂlecÂtuÂal capÂiÂtal [must be] treatÂed as manÂagers in their own right, and posÂsess the exÂecÂuÂtive auÂthorÂiÂty reÂquired to deÂcide how best to deÂploy their time.
I think that the future belongs to those who will be able to quickly master new things. The so-called specialists will face mounting pressure as new non-specialists start to occupy their territory with greater and greater ease.
Unless you manage, you do must have a certain specialization. Managing is a different issue, the technical knowledge does not matter much. What matters is the ability to bullshit confidently.
Daniel,
I think we define generalization too differently. Actually, I was working in a small team that was delivering a web-based e-mail to millions of people so that I could see what people were doing. Even those who were doing a lot of different tasks, in my opinion, specialized in a very narrow field. That is they had experience with C++, Linux, SQL, perl, mail protocols, and some search algorithms. If they were generalists they would also write phone aps, design IDE for Windows, MAC, etc…
@Itman
Take the best in this this team you describe. How long would it take them to master phone app programming and design?
Now, let’s compare with how things were 20 years ago. How long did it take for someone to master a new platform, a new type of applications?
My point is that it is getting easier, not harder, to catch up to the experts in any field.
@Itman
I stand by my statement:
Your future wealth is determined by how much you can expand your mind beyond the capacity of your biological brain, not by your current skills.
There is nothing risky about this proposition.
It is definitely easier. It is just not feasible unless you manage. I am afraid of the perspective when everybody is becoming a super-generalist (without deep technical expertise) and who is doing the business? I want a plane to fly smoothly instead of crashing. Just because somebody was catching up with the experts too quickly.
Daniel, I completely agree with you on the statement. Your or mine wealth does depend on it. However, this model is not scalable. If everybody starts to generalize, the civilization will collapse. Well, and anyways if I, e.g., don’t like overgeneralization, I would prefer to stick to my technical skills and be happier despite it would mean less money.
An example of generalization inside Google?
http://www.smashingmagazine.com/2011/07/08/interview-google-marcin-wichary/
I rather enjoyed reading your perspective, particularly when you’ve taken the time to reason about it so.
I would enjoy reading more of your perspective on the topic in the future.
– Bill
@gregbo
(…) questions asked required specific knowledge such as what TCP port a service ran on or what tcpdump command was used to print certain types of packets.
Ironically, you can google these questions very easily.
These are not questions I would ask a potential new Google engineer.
Hmmm. I didn’t see much evidence of generalization in the Marcin
Wichary article. The subject of the article is primarily a user
experience (UX) engineer who leverages various web technologies
(e.g. HTML5). I didn’t see much evidence of cross-disciplinary
involvement, even within software engineering (e.g. operating systems
or AI).
My experience tends to match Itman’s, especially with regard to
Google. The interviews I’ve had with them have been very specific and
detailed; questions asked required specific knowledge such as what TCP
port a service ran on or what tcpdump command was used to print
certain types of packets. Not the higher-level conceptual engineering
types of questions I was expecting. On a more general note, most
people are not capable of being outstanding in several subfields of
the computer industry, so they are forced to specialize in order to
pass an interview such as the Google interviews I’ve had. Managers
benefit from generalization more, because it enables them to employ
multiple strategies to run their teams, but individual contributors seem
to be judged on highly specific topics such as how to code a particular
algorithm, etc.
Perhaps Google has changed their interview strategies, but I haven’t
seen much evidence of it from other sources of information on Google
interviews.
Daniel, it probably depends on a department. The last time I talked to a Google recruiter, ironically, they said they wanted a generalist. What they probably meant, however, was that the person would know C++, Unix, network protocols, databases, and a bit of the WEB (as opposed to, e.g., SQL coder), rather than being a computer scientist, a successful coder proficient in Linux/Windows/Mac technologies, who is running his own business in spare time.
Even when technology enables easier access to knowledge doesn’t mean there wont be hyperspecialization. Even if we are expanding our minds exponentially, the knowledge base in the world is expanding even more exponentially! Because of this increasing distance between the general knowledge base and our brains ability to handle information, you have to focus on a narrower area in order to do any meaningful contribution to this knowledge base.
Not saying that generalists doesn’t have their place too, but more in the role of creating applications with existing knowledge. So I wouldn’t say it’s more risky to work as a generalist than specialist.