<p>
</p>
<p>Or, I suppose, betting your undergraduate study on landing a high-paying job in consulting or finance. Interestingly, many of the better performing Berkeley engineering students that I know would not even take engineering jobs at all, but would instead take jobs in consulting or finance. Higher performing liberal arts students can do and have done the same. </p>
<p>The other strategy is to develop marketable skills on your own time. By choosing a creampuff major, you can earn strong grades with little studying, hence enjoying vast quantities of spare time that you can devote towards developing marketable skills outside of the classroom. For example, I know a guy who majored in a humanity at Berkeley, while also learning Web programming in his spare time, which he supplemented with part-time jobs building websites and web applications. By the time of graduation, he was a savvy web programmer with an impressive portfolio of prior work, and was actually garnering higher-paying job offers than many of the actual CS/EECS graduates. {As to why he didn’t just major in CS/EECS, his interest in programming wasn’t sparked until he was already well into his major by which time it was a bit late to switch majors, and, frankly, he was also repulsed by the CS grading curves. Furthermore, the CS program surprisingly teaches relatively little Web programming and provides few opportunities to develop a portfolio of practical projects.} He’s now an award-winning software developer at his employer, despite not actually holding a CS degree. </p>
<p>I have similarly recommended that more Berkeley students should become IT workers. The vast majority of IT workers, at least that I know, were not talented enough to be admitted to high ranked colleges and do not even hold technical degrees, or often times hold no college degree at all (with a few never even graduating from high school). Yet they make highly respectable salaries encompassing well over 6 figures with sufficient experience. A typical Berkeley student could probably become a Cisco certified CCIE with 3-6 months of full-time study or 1-2 years of part-time study, and that extra time would be readily available via a creampuff major that places few demands on one’s time. </p>
<p>One could also even develop practical skills within the Berkeley classroom, but through a P/NP basis. For example, one of the most promising courses I have ever seen is [Statistics 133](<a href=“http://www.stat.berkeley.edu/classes/s133/resources.html”>http://www.stat.berkeley.edu/classes/s133/resources.html</a>) which, rather than bogging students down in (unmarketable) theoretical statistical formula, instead teaches students how to obtain, parse, organize, and mine actual real-world datasets and then present results, using actual software that is widely utilized within industry. You will learn R, surely the most widely utilized open-source statistical software standard in the world. {And once you’re versed in R, then learning a commercial stat package such as Stata or SAS is relatively simple.} You will learn MySQL, Tcl, and CGI - technologies that are widely utilized throughout industry to organize and manipulate databases. Perhaps most importantly- the class has no prereqs. Anybody, even a humanities student with no prior computing or statistical training, could take this course. Those who are worried about the grading could simply take it on a P/NP basis, or heck, even sit in on it without actually taking it. Frankly, I think everybody should take or at least sit in on this course, and Berkeley should vastly increase the number of seats for the course. </p>
<p>Now, to be sure, I obviously don’t claim that you will develop sufficiently marketable skills to land a high-paying job via a single course alone. However, the course will provide you with a foundation upon which you can build those skills. The hardest part of learning any software skillset is the starting point; once you’ve learned enough to implement basic functionality, it’s not particularly difficult to develop additional skills that builds upon that foundation. For example, once you’ve learned how to use R to import, manipulate, and calculate basic statistical regressions on simple datasets, it’s not that hard to learn how to analyze more complex data. </p>
<p>One could then take on research projects that rely solely on statistical data analysis, perhaps as part of a widely praised undergraduate thesis that has drawn nationwide media [attention</a>](<a href=“http://blogs.wsj.com/deals/2010/03/15/michael-lewiss-the-big-short-read-the-harvard-thesis-instead/]attention”>http://blogs.wsj.com/deals/2010/03/15/michael-lewiss-the-big-short-read-the-harvard-thesis-instead/) and has now been placed on the reading list of an economics course. One could perform a fun statistical analysis on topics of personal interest, and even have it published in a peer-reviewed journal, such as a statistical analysis of the social network of [url=<a href=“http://arxiv.org/abs/physics/0511215]rappers[/url”>[physics/0511215] The Network of Collaboration Among Rappers and its Community Structure]rappers[/url</a>]. {Apparently, Snoop Dogg, Busta Rhymes, and 2Pac are heavily ‘connected’ rappers.} Or one could simply find a job that requires statistical data analysis skills - nowadays companies are swimming in data and just don’t know what to do with it. {Heck, the supposed value-add of most consulting firms is usually to provide that analysis.} </p>
<p>The upshot is that plenty of avenues exist for students in creampuff majors to develop highly marketable skills outside the classroom - or even through judicious choices within the classroom. One merely has to make wise choices with the large dollops of spare time that the creampuff major provides. The ‘ingenious’ aspect of the strategy is that you can do so without ever having to jeopardize your GPA. For example, if you attempt to learn Web programming or Cisco networking skills in your spare time yet perform poorly - for example, if you earn the worst score in the history of the CCIE certification exam - nobody will ever know about it. Indeed, most CCIE’s fail the certification exams multiple times before they finally pass, but that’s harmless because evidence of your failures is not recorded. In contrast, if you take engineering courses and earn poor grades, those poor grades are indelibly marked on your academic record for life.</p>