“We knew Jian-Yang’s app was in a food space, yet we insincere that it was camera-based,” says a indistinguishable try entrepreneur in a new part of HBO’s Silicon Valley. “Like we take a print of food, a app earnings nutritive information or recipes or how it was sourced.”
Jian-Yang, one of a programmers who live during a “incubator” in that a uncover is set, has usually pitched an app called “Seafood” to VCs during a illusory organisation Coleman Blair. His thought is to foster his grandmother’s octopus recipes (she gave him a family recipe, he says, “before she died in a terrible way”), yet Seafood fast gets disfigured into a representation Coleman Blair expected.
By a finish of a meeting, Jian-Yang’s app is called “See Food,” it’s a “Shazam for food,” and Coleman Blair has invested–even yet a record doesn’t indeed exist. (Shazam is an app that can brand songs from usually brief clips.)
The whole stage is ridiculous. It’s also not so opposite from a genuine app that launched in 2011.
Called “Meal Snap,” a app took users’ food photos, identified a dishes within them, and returned their calorie essence in real-time. It, too, worked like a “Shazam for food.”
Daily Burn, a association that combined Meal Snap, ran a whole apartment of fitness- and food-tracking apps during a time, yet this sorcery food marker record seemed special. “With a tiny some-more speed and accuracy, Meal Snap could join a pantheon of truly jaw-dropping apps,” CNET wrote during a time.
Like See Food, however, Meal Snap wasn’t wholly upfront about a tangible state of a technology.
In a HBO-version of Silicon Valley, Jian-Yang and Erlich Bachman, who runs a incubator, eventually come adult with an programmed food tagging system. The problem is that it can usually brand “hot dog” and “not prohibited dog,” and identifying some-more dishes will need scraping unconstrained images of food from a internet to use as training information for a computational indication (a charge Erlich attempts to pretence a Stanford mechanism scholarship category into completing for free).
Back in 2011, as Meal Snap launched, identifying food in genuine time regulating synthetic comprehension was out of a question, even if we could partisan a overflow of intelligent millennials to help.
Meal Snap’s website, referred to a record as “magic.” So did Daily Burn CEO Andy Smith, in interviews. Though he concurred that a app’s sorcery concerned humans, he never offering specifics about how it worked.
Here’s how it worked: After users took a photo, Daily Burn routed it to Mechanical Turk, an Amazon pursuit marketplace where companies can compensate workers cents to do tiny tasks like brand photos. Mechanical Turk workers identified a photos as “apple” or “chicken” or “burger,” and Daily Burn matched their descriptions with a database of opposite foods’ calorie contents.
Daily Burn pivoted shortly after (the Meal Snap app in sold wasn’t a good business, as Mechanical Turk fees piled adult each time a user snapped a photo). It now streams online aptness classes, that it sells entrance to on a subscription basis.
“I don’t watch Silicon Valley,” Smith told me when we reached him today. “It’s too tighten to home. It stresses me out.”