Monday, August 17, 2015

America needs more H1B visas, but (probably) won't get them

The current US political climate is increasingly anti-immigration, including high-skilled immigration. This not only makes much-needed reforms of the H1B visa system increasingly unlikely, but suggests the program might be considerably scaled back. Unfortunately, I've been dealing with H1B-induced annoyances my entire career so far, and it looks to continue. The latest: my attempt to hire an internal transfer at Microsoft was stymied because the change in position would reset their H1B visa application. Note this is someone who already is in the United States and already works at Microsoft.

So clearly immigration laws are not designed to optimize either allocation efficiency or human welfare. However, perhaps there is a more cold-hearted calculation in favor of the current regime? I don't think so.

Economic Nationalism. If the point of immigration laws is to make America richer, it's a fail. With technology, a laborer can create value anywhere with (intermittent!) electricity and internet. All the immigration restrictions have done is teach companies how to acquire talent in their home markets. Not only does America lose out on the direct tax revenue, but also secondary economic activity such as demand for housing, infrastructure, transportation, education, entertainment, child care, etc. Case in point: check out Microsoft's increasing footprint in Vancouver, where immigration laws are more sane. Funny side note: collaboration with employees in the Vancouver office is made more complicated by immigration laws, e.g., they cannot visit on-site in Redmond too frequently. Three (Bronx) cheers for regulation.

Protecting American Workers. Ok, maybe these regulations don't help America at large, but do benefit domestic technology workers. I don't buy it, because the resulting reduction in labor's bargaining power degrades the quality of the workplace. Let me explain. Technology workers who have not obtained a green card have two very strange properties: first, they have a large amount of non-monetary compensation (in the form of legal assistance with the green card process); and second, they have limited freedom to change their job during the visa process. These two effects combine to greatly reduce the bargaining power of foreign technology workers, who in turn are willing to accept less money and worse working conditions. Consequently, domestic workers have their collective leverage over employers reduced because part of the labor pool is unable to negotiate effectively. If visa restrictions were relaxed, labor conditions for domestic and foreign employees would both improve.

Promoting Innovation. Another fail for our current policies. I spent the first half of my career in startups, where everyone has at least a green card if not a passport. No one in the visa process can afford the inherent volatility of a startup (side note: kudos to Halt and Catch Fire for converting “missing payroll” into great television). The net result is that startups are starved for human capital disproportionately to large firms, as the latter have the capital and expertise to both navigate the legal process and engage directly in overseas labor markets. Favoring incumbents over insurgents? Not exactly a formula for creative destruction.

To summarize: I'm very unhappy with the current mood of the American electorate. It's not just mean, it's also bad for the country.

By the way, if you are looking for a job, please contact me as indicated in the top-right position of my blog. My blog has been continuously advertising open positions where I work since I started it, because my entire career I have always worked on teams with open positions that go unfilled. Funny that.

Monday, August 10, 2015

Paper Reviews vs. Code Reviews

Because I'm experiencing the NIPS submission process right now, the contrast with the ICLR submission process is salient. The NIPS submission process is a more traditional process, in which first (anonymous) submissions are sent to (anonymous) reviewers who provide feedback, and then authors have a chance to respond to the feedback. The ICLR submission process is more fluid: non-anonymous submissions are sent to anonymous reviewers who provide feedback, and then authors and reviewers enter into a cycle where the author updates the arxiv submission and reviewers provide further feedback. (ICLR also has public reviews, but I'm not going to talk about those). Note in the traditional model the reviewers have to imagine what my (promised) changes will look like in the final version.

The traditional model is from an age where papers were actual physical objects that were sent (via snail mail!) to reviewers who marked them up with ink pens, and hopefully advances in technology allow us to develop a more effective process. I think we should look to software engineering for inspiration. Having worked both as a researcher and as a software engineer, I appreciate the exoskeleton-robot distinction. In this context, the exoskeleton posture of science lends itself to a ballistic concept of paper review, where a completed unit of work is accepted or rejected; engineering is more about collaborative continuous improvement. Truthfully, most journals and some conferences utilize a more fluid review process, where there are “conditional accepts” (changes need to be re-reviewed) and “shepherds” (reviewers who commit to guiding a paper through several rounds of reviews). These processes place more burden on the reviewers, who are providing the valuable service of helping someone improve their work, without compensation or recognition. Naturally conferences might be hesitant to demand this from their volunteer reviewer pool.

The solution is technology that eases the cognitive and logistical burdens on all parties. Code reviews have the same broad goal as paper reviews: improvement of quality via peer feedback. Here are some things we can learn from code reviews:
  1. Incremental review. Rarely a programmer develop a complicated piece of software de novo. Most reviews are about relatively small changes to large pieces of software. To ease the cognitive burden on the reviewer, the changes are reviewed, rather than the entire new version. Visualization technology is used to improve change review productivity.
    1. The initial submission of a paper is distinct from the typical code review in this respect, but the subsequent cycle of revisions aligns with this nicely.
  2. Modular updates. When a programmer makes several distinct changes to a piece of software, the changes are (to the extent possible) designed to be commutable and independently reviewed. Technology is used to facilitate the composition of (the accepted subset of) changes.
    1. This aligns nicely with the review process, as different reviewers' feedback are analogous to issues.
  3. Minimal clean changes. Smart programmers will craft their changes to be most intelligible under review. This means little “easy” things like avoiding lexical changes which are semantically equivalent. It also means a tension between cleaning up the control flow of a program and creating a large change.

Add to this the ability to preserve the anonymity of all parties involved, and you would have a pretty sweet paper review platform that would plausibly accelerate the pace of all scientific fields while improving quality.

This is precisely the kind of public good that the government should fund. Somebody in academia: write a grant!