It’s great to hear that Twitter is relying on machine learning to identify spammy accounts, and it particularly makes me glad to hear that they are relying on more numeric indicators (e.g., number of connections, which users they are connected to) as markers for identification, rather than natural language processing of the tweets themselves which would be subject to far more bias. However, as pointed out by a commenter above, the 80% accuracy rate is concerning. That’s not to say that every algorithm should be expected to be totally accurate, but it does make me wonder what kind of input is inserted into the process of evaluation alongside the algorithm for identifying and removing spam accounts.
I really appreciated this view into the manner in which another clothes retailer is pairing data science with trendsetting. However, as we saw in the case with Gap, I worry whether this total reliance on “data” to spot trends and tailor them to individuals is going to be actually effective. While a comparison to StitchFix is made by the author, it’s important to know that StitchFix’s algorithm and selection takes into account the fact customers have different tastes, and the algorithm is used in tandem with “stylists” who rely on a combination of algorithm outputs and their own judgement make the final recommendation on fashion pieces. Based on the description above, it almost sounds as if ZOZO is relying on a one-size-fits-all machine learning algorithm, which leads me to worry whether it will be as effective in understanding the diversity of trends consumers want.
I really appreciate this case example of where open innovation might be considered a “failure”, but I wonder if this failure is specific to conditions that are unique to the process of entertainment production. First, when considering the huge volume of individuals (from average joes to professional writers) who have ideas for scripts and shows, I worry whether the much wider production funnel of having to process through suggestions would have been a feasibly surmountable challenge for really any group or organization, even with crowdsourced prioritization. Second, even if the number of suggestions were somehow made to be more manageable, I’d question whether crowdsourced selection is truly what a creative industry like entertainment production should rely on. There’s a reason the entertainment industry is led by creatives and artists: they have an ability to have an artistic vision and see ahead to the “next thing” that will provide artistic value to viewers, and I’d argue there is an artistic intuition required for such an eye for selection that the masses won’t have.
Honestly, this type of open innovation campaign from PepsiCo is brilliant from a cost-savings perspective, for two reasons. The first is the fact that PepsiCo has essentially “outsourced” its R&D and marketing research to the masses, saving money on otherwise expensive in-depth market research activities. The second is that they have essentially turned this into a PR/marketing opportunity, thereby saving costs from what would otherwise be a “just another” release of another PepsiCo product. It’s interesting to think of how other companies in the future similarly find cost-saving synergies between open innovation initiatives and other business-as-usual activities.
The question around scaling this technology that the author poses is the same question that sits at the front of my mind upon reading this article. In reading through, I immediately thought of the potential for this technology to replace (or at least, mimic at a commodity-goods level) the types of creative chocolate sculptures and goods usually produced by the most talented artisan chocalatiers and chocolate artists. However, in the same way those artisans’ skillsets aren’t easily replicated, I wonder whether the printing technology at the Hershey’s Chocolate World can be replicated to fill Hershey’s production facilities and factories.
Such a fantastic view into the intersection of addititive manufacturing & the automotive manufacturing industry, particularly at a trendsetting company like BMW. To be frank, most discussion of “3D printing” conjures images of smaller printers not used to produce items at the size scale of an automobile. My biggest questions after reading through this though are twofold. First, what are some very specific product innovations that BMW plans on attempting with additive manufacturing (e.g., is it a detail like a spoiler, or an entire car body)? Second, who are the suppliers building the additive manufacturing equipment that BMW is using to fill its warehouse, and will that technology ever develop to be able to produce automobiles at the same speed & scale that manufacturers like BMW currently produce their goods?