Elisa Vandensteene

No ratings yet.

In 2017, a few schools started accepting cryptocurrency tuition fee payments, on the grounds that the transaction fees and speed made it a convenient alternative to checks or wire transfers. In New York, the Montessori preschools accept Bitcoin or Ethereum but no credit cards, arguing that credit card fees are two, three, or even four times higher than Coinbase fees, their digital currency exchange. Additionally, credit cards payments often get declined, which leads to more fees and takes even more time. (Dangremund, 2017)

While it may by now be mainstream for tech businesses to accept cryptocurrency payments, some can find it surprising that non-digital institutions, such as schools, are accepting this payment option. What would be the benefits and downsides for traditional businesses to implement it? How should they assess and manage risks?

Benefits

As previously mentioned, cryptocurrency transactions are generally much cheaper and faster than more traditional options. Companies who adopt this payment option may broaden their customer base, as offering more payment options is linked to higher conversion rates. A reason for this is that to pay in cryptocurrencies people only need to have access to the internet, no bank account is necessary, so the problem of some customers having foreign accounts would be resolved. Additionally, as blockchain is global, they are also not bound by exchange rates. It can help attract a younger demographic, people who prefer the simplicity and anonymity of blockchain. Lastly, for businesses that sell pricier goods or services, another benefit may be that it is a good medium for crowdsourcing, meaning that several people can contribute to a single purchase, as everything goes through a trustworthy smart public ledger. (Harrison, 2018)

Downsides

The main downside of accepting crypto payments is the higher risk exposure. If a traditional company wants to implement this, it will most likely do it through crypto exchanges, which act as processors for businesses that do not directly trade in cryptocurrencies (Harrison, 2018). Exchanges entail transaction fees, liquidity concerns, and counterparty risks (Harrison, 2018). Leaving money in exchanges exposes companies to volatility, as well as to theft vulnerability (Harrison, 2018). However, this risk can be easily managed by either directly converting crypto payments to normal currencies, or by having a hard wallet offline, for instance on Ledger Nano (Harrison, 2018). Another, rather unrelated, downside to blockchain is the environmental impact. Bitcoin is extremely energy intensive, which is not exactly a good characteristic to have considering the current climate crisis (Irfan, 2019). While a lot of consumers may not care about it, others may be driven away from companies who accept Bitcoin, due to the fact that they are promoting unsustainable practices.

References:

Dangremond, S. (2017). Why These New York City Private Schools Are Accepting Bitcoin. Retrieved 9 October 2019, from https://www.townandcountrymag.com/society/money-and-power/a10207209/montessori-schools-bitcoin/

Harrison, K. (2018). Should Your Company Accept Bitcoin And Other Cryptocurrency Payments?. Retrieved 9 October 2019, from https://www.forbes.com/sites/kateharrison/2018/09/10/should-your-company-accept-bitcoin-and-other-cryptocurrency-payments/#5904d14f3373

Irfan, U. (2019). Bitcoin is an energy hog. Where is all that electricity coming from?. Retrieved 9 October 2019, from https://www.vox.com/2019/6/18/18642645/bitcoin-energy-price-renewable-china

5/5 (4)

Data based decision making is widespread across industries and departments, therefore impacting resource allocation, product development and design choices, human resource management practices, and so forth. So what if data is only seemingly objective, but actually inherently biased?

The result is that, in spite of best efforts to decrease gender and race inequalities, biases are not only perpetuated, but amplified. These biases lead to women and minorities being treated unfairly, as well as put their lives in danger. The issue is that oftentimes the “default white man” is treated as the “default human” in data analysis, and therefore decision making. This happens, for instance, when healthcare devices are designed according to male physical characteristics, such as the electrical wave threshold below which a pacemaker is fitted being correct for men. Unsurprisingly, it is therefore not well suited for women, although sold for both genders. Or worse, when clinical trials are not sex-disaggregated, and high blood pressure drugs end up reducing the chance of heart attacks for men, but increasing it for women. (Gordon, 2019)

These kinds of issues are likely to become more important in the future, when the majority of processes will be automated, thus unsupervised, and driven by machine learning. A machine learning algorithm needs assumptions, and relies on statistical bias to make predictions (Schadowen, 2018). This becomes dangerous when an algorithm is trained on gender or race biased data, and faulty assumptions are incorporated in the model, resulting in machine bias. Contrary to statistical bias, machine bias is not necessary to make predictions, but rather the result of prejudice being assumed by the model, from either its creator or training data (Schadowen, 2018). Therefore, it is possible to rid systems of machine biases, but it requires that developers acknowledge them, and account for gender and race differences while developing, testing, and re-evaluating their models. This means that counterintuitively, to achieve equality it is actually necessary to treat men, women, and minorities as unequal, account for physiological and lifestyle differences, and adjust the models accordingly.

References:

Gordon, S. (2019). It’s a man’s world — how data are rife with insidious sexism | Financial Times. Retrieved 18 September 2019, from https://www.ft.com/content/9e67294a-28a0-11e9-a5ab-ff8ef2b976c7

Marriott, J. (2019). Data Discrimination: Exploring Big Data and Bias|SXSW 2015 Event Schedule. Retrieved 18 September 2019, from https://schedule.sxsw.com/2015/events/event_IAP43058

Shadowen, N. (2018). How to Prevent Bias in Machine Learning. Retrieved 18 September 2019, from https://becominghuman.ai/how-to-prevent-bias-in-machine-learning-fbd9adf1198

Should non-digital businesses accept Bitcoin payments?

9

Please rate this

Is machine learning hard-coding discrimination in the world?

18

Please rate this