The comment that stood out to me in your post was “machine-learning relies on adequate and accurate input”. I agree with this statement and it has made me skeptical about some of the practices that Schlumberger is emplying for its data lake. Collaboration across Schlumberger customers and partners is a great way to get a large data set, but I wonder if data coming from mutliple sources might inherently have some bias. It is possible that some customers might have over representation in the data set which may add bias towards a specific practice or geography. Collaboration across customers can also be difficult because there are sometimes additional context or factors that may be confidential and are not included in the shared data causing an impact on the validity of the model.
The two most interesting point that you brought up were the reduction of waste with AM and the possibility of parts-on-demand in the future. For GE this seems like the most logical way to progress with this technology. The reduction of waste would fit with GE’s sustainability initiatives, since for many industries AM can signficantly reduce waste products during the manufacturing process. Instituting parts-on-demand may require GE to collaborate with their suppliers and increase training programs, not only for their internal engineers, but for their customers technicians. GE should focus on pitching these benefits to their customer and partner closely with them during their proof-of-concept process.
While reading your commentary on Visa’s future applications for machine learning, I kept thinking about user security. Metrics such as mobile phone location and especially biomarkers of users, made me question how much privacy I would be willing to give up in order to increase my credit card security. This may be a difficult balance for Visa. I also agree with your hesitancy that a machine learning model may eventually be used as a tool or guideline to commit undetectable fraud. With both of these in mind, Visa will have to be careful with how it approaches this problem and should have its own security measures in place to ensure that this is successful.
The Foundry seems like an interesting solution to a problem that is plaguing the CPG industry. I wonder whether this will be the long term solution to Unilever’s problems with market decline or if this is just a band-aid solution that is delaying the inevitable. I agree with your assessment on making long term partnerships to help them stay innovative, as opposed to one off projects. As a large consumer goods company, I would be primarily concerned with how to stay innovative long term either from an aquisition of a innovative small company or by changing the culture within. If they do not make any changes, I could see a well-funded creative and innovative company stealing more market share in the future.
This article leaves me wondering what kind of bias a company might face when using open innovation and how reliable that information will be about their customer base. For the Vetr example you used, I can imagine that the crowd sourced data set might be overly bias towards the technology sector due to the nature of it being a web-based platform. For the Barclaycard Ring Mastercard, I would question if the voting mechanism accurately represents the cardholder’s interests. There are ways to ensure that open innovation is coming from the target customer. The best example of this done right is the one you provided about gaming mods. This is a bottom up approach driven by early adopters and high frequency gamers that drive the industry trends. If Barclays can ensure that their open innovation data is not biased and are aware of the specific customer base that is contributing the majority of the information, then I agree that it will be a competitive advantage for the bank.
I would question the hesitancy of NASA to move forward with AM for mission critical components. I agree that there are many unknowns and high risks associated with this practice, but it would benefit NASA to prioritize studying and derisking these unknowns. The use of AM for mission critical components would have several key benefits, primarily reducing cost of space shuttles. The Space Shuttle Endeavour cost NASA, and therefore tax payers, approximately $1.7 billion . Through the use of cheaper, but reliable, parts from AM could have a significant impact about the amount of government spending on space shuttles and could potentially increase scientific discovery by sending more lower cost shuttles in space.
 NASA, “Space Shuttle and International Space Station” URL: https://www.nasa.gov/centers/kennedy/about/information/shuttle_faq.html#1 [Accessed 15 November 2018]