Ａ Ｓｅａｔ ａｔ ｔｈｅ Ｄａｔａ Ｔａｂｌｅ .
Article by @urban_ekonomics .
Full article at urbanEkonomics.org
...Humans are full of cognitive biases, many of which are unconscious but none the less influence our decision making. The summation of these biases are captured best with one of my favorite biases, the “blind spot bias.” 𝕋𝕙𝕖 𝕥𝕖𝕟𝕕𝕖𝕟𝕔𝕪 𝕥𝕠 𝕤𝕖𝕖 𝕠𝕟𝕖𝕤𝕖𝕝𝕗 𝕒𝕤 𝕝𝕖𝕤𝕤 𝕓𝕚𝕒𝕤𝕖𝕕 𝕥𝕙𝕒𝕟 𝕠𝕥𝕙𝕖𝕣 𝕡𝕖𝕠𝕡𝕝𝕖, 𝕠𝕣 𝕥𝕠 𝕓𝕖 𝕒𝕓𝕝𝕖 𝕥𝕠 𝕚𝕕𝕖𝕟𝕥𝕚𝕗𝕪 𝕞𝕠𝕣𝕖 𝕔𝕠𝕘𝕟𝕚𝕥𝕚𝕧𝕖 𝕓𝕚𝕒𝕤𝕖𝕤 𝕚𝕟 𝕠𝕥𝕙𝕖𝕣𝕤 𝕥𝕙𝕒𝕟 𝕚𝕟 𝕠𝕟𝕖𝕤𝕖𝕝𝕗
Now my argument does not suggest that the data scientists behind these algorithms are inherently evil or lack compassion. I suggest simply that they are human, comprised of their lived experiences and plagued by biases.
In fact most data scientists would wholly agree with the now famous quote. “ᴀʟʟ ᴍᴏᴅᴇʟs ᴀʀᴇ ᴡʀᴏɴɢ, ʙᴜᴛ sᴏᴍᴇ ᴀʀᴇ ᴜsᴇfᴜʟ”
Models are comprised of an acceptable error range, and the predictability of these models sometimes create value.
“Data scientist” in it’s singular form is a misnomer, as it’s most often a team comprised of business practitioners, statisticians and technologists. A team with a collective set of beliefs, lived experiences and biases.
Making decisions with far greater implications than backlash from a poorly constructed Pepsi ad.
As the application of big data grows, where is the conversation around representation within these data science teams? A field that has real world implications for large portions of the population.
The most obvious response is that there aren’t enough “data scientist” from under represented communities and the pipeline is sparse. That these education programs lack funding and students would rather pursue careers in less technical fields.
Well conversation can spark targeted action, and that targeted action can help us avoid bringing a new phrase to infamy. “ᴀʟʟ ᴍᴏᴅᴇʟs ᴀʀᴇ ᴡʀᴏɴɢ, sᴏᴍᴇ ᴀʀᴇ ᴅᴇᴛʀɪᴍᴇɴᴛᴀʟ” -Urban Ekonomics
One solution, in its 3rd year is a scholarship aiming to get Chicago Public School students to consider #technology Visit urbanekonomics.org for more #blackintech #afrotech