CS2125 Paper Review Form - Winter 2019

Reviewer: Ali Harakeh

Paper Title: Machine learning, social
learning and the governance
of self-driving cars.

Author(s): Jack Stilgoe

1) Is the paper technically correct?
 [x] Yes
 [ ] Mostly (minor flaws, but mostly solid)
 [ ] No

2) Originality
 [ ] Very good (very novel, trailblazing work)
 [x] Good 
 [ ] Marginal (very incremental)
 [ ] Poor (little or nothing that is new)

3) Technical Depth
 [ ] Very good (comparable to best conference papers)
 [ ] Good (comparable to typical conference papers)
 [x] Marginal depth
 [ ] Little or no depth

4) Impact/Significance
 [ ] Very significant
 [x] Significant
 [ ] Marginal significance.
 [ ] Little or no significance.

5) Presentation
 [ ] Very well written
 [ ] Generally well written
 [x] Readable
 [ ] Needs considerable work
 [ ] Unacceptably bad

6) Overall Rating
 [ ] Strong accept (award quality)
 [ ] Accept (high quality - would argue for acceptance)
 [x] Weak Accept (borderline, but lean towards acceptance)
 [ ] Weak Reject (not sure why this paper was published)


7) Summary of the paper's main contribution and rationale
   for your recommendation. (1-2 paragraphs)

Disclaimer: I am not an expert on the topic studied by this paper, and this review should be read while taking that into consideration.

This paper argues that trajectories and rhetoric of machine learning in transport pose a substantial governance challenge. 
The paper sheds light on the fact that most current players in self-driving technology are forced to exaggerate the speed at which they can push their product to the consumers’ hands. As a result, 
companies are shipping systems that work well 999 times of 1000 and then catastrophically fail leading to human causalities. The paper goes on to present the vast differences in approaches between top players, 
and the lack of education of what the public perceives as an “autonomous vehicle”. The paper then suggest that democratizing learning might be a good approach to reduce these problems. Democratizing learning 
is twofold, by educating the public on the limitations of self-driving technology, and through legislations that push corporates to share the enormous amount of private data they are currently hording. The author 
urges legislators to have such conditions before allowing for public testing of autonomous vehicles. 


8) List 1-3 strengths of the paper.  (1-2 sentences each,
identified as S1, S2, S3.)

S1: The paper sheds light on how the collaboration between top players has a higher public value than the current state of competition, especially in making autonomous vehicles safer.

S2: The paper sheds light on how misinformation provided by companies regarding their technologies leads to catastrophic accidents.


9) List 1-3 weaknesses of the paper (1-2 sentences each,
identified as W1, W2, W3.)

W1: The paper shows some bias against competition and is formulated as more of an opinion piece than a truly objective one. Competition is what drives innovation in autonomous driving, 
and the full removal of this aspect will substantially hinder progress of the field in my opinion.

W2: The paper focuses on Tesla as a running example. It would have been also beneficial to look at how other top players are handling product advertisement and safety related issues.

W3: The paper's strucure is a little bit confusing. I could not form a relation between sections, but that might just be my inexperience in reviewing non-technical papers.