3 ms·
It's fairly common to see an element of outright fakery before the hopeful-but-not-guaranteed "till you make it". A coworker was using a mobile app from a star
by jiggawatts 5y ago
It's fairly common to see an element of outright fakery before the hopeful-but-not-guaranteed "till you make it".
A coworker was using a mobile app from a startup offering to do "receipt OCR using AI/ML" when in fact they were sending most of the scanned pictures to a call-center in India to be manually entered. They hadn't yet figured out the fancy image recognition algorithm, so they were just throwing AWS Mechanical Turk at it while building up a user base. Realistically, they might never get to that level of AI while staying solvent.
That's not so different ethically to what Theranos was doing. The only difference with Theranos is the scale of it all.
- rocqua 5y agoThere are two parts to this comparison. The first is, do they outright lie. If they are claiming "we are using AI" to investors, that might be fraud. But that is avoidable with just adding in "we are using AI for some receipts". The second, and more important, question is, are they harming customers. Theranus gave people unreliable blood test results and lied about their reliability. That does actual harm to customers. Whereas for scanning receipts the only thing a customer really cares about is accuracy.
- josho 5y ago> That's not so different ethically to what Theranos was doing No. One is we know we can do OCR, we know we can do AI, it's just an engineering effort to put those pieces together. The risk is schedule risk. Theranos was hard science. They did not know if they could do what they fraudulently claimed they had already done. They needed scientific discoveries to deliver. The risk was can the necessary technology be invented.