Shulou(Shulou.com)06/02 Report--

2019-09-14 11:47:26

Kiwibot, an online celebrity food delivery unmanned car at the University of California, Berkeley.

After being kidnapped and waxed by students for spontaneous combustion, it got into trouble again, and it was severely criticized:

Although it claims to be autopilot, it is actually operated remotely in Colombia-up to three robots per person for less than $2 an hour, higher than the local minimum wage.

Employing people to pretend to be AI, a story full of "fraud", is easy to attract widespread condemnation. But for any company with artificial intelligence as the label, this can be called "heart-to-heart theory".

Not long ago, a company called Engineer.ai attracted a lot of attention when it was accused of impersonating a programmer from India as AI.

Now that the waves are rising again, it is more worthwhile to reflect: how on earth should we treat this accusation? And when blaming them, what on earth should we blame?

Real artificial, fake AI?

Kiwibot, from a startup called Kiwi Campus, was founded at the University of Berkeley in 2017.

The motivation was that founder Felipe Chavez and others found that ordering takeout on American campuses was as expensive as pizza, so they wanted to develop a robot to reduce the cost of food delivery.

At first, Kiwibot delivered food only on the Berkeley campus, and later expanded to other schools, such as Stanford. According to its website, the company already has more than 150 robots and has completed more than 30000 delivery services.

And it has become a hot topic on social networks from time to time:

In December, it ignited spontaneously at the University of Berkeley because the battery was not properly replaced by staff, prompting students to light wax to mourn.

In May this year, because some people did not like it, they simply "kidnapped" it and put it in the trunk.

So far, Kiwi Campus has received seven rounds of "seed wheel" financing, totaling $2 million, or about 14.3 million yuan.

Initially, Kiwibot was accused of hiring a manual to operate a food delivery car, according to a report in the San Francisco Chronicle:

The Kiwibots do not figure out their own routes. Instead, people in Colombia, the home country of Chavez and his two co-founders, plot "waypoints" for the bots to follow, sending them instructions every five to 10 seconds on where to go.

Kiwibots does not calculate its own route. Instead, people in Columbia, the hometown of Chavez and his two co-founders, set a "path point" for robots, sending them instructions every five to 10 seconds to tell them where to go.

As with other offshoring arrangements, the labor savings are huge. The Colombia workers, who can each handle up to three robots, make less than $2 an hour, which is above the local minimum wage.

Like other offshore outsourcing, it saves a lot of labor. Workers in Colombia can operate up to three robots for less than $2 an hour, higher than the local minimum wage.

Report link:

Https://www.sfchronicle.com/business/article/Kiwibots-win-fans-at-UC-Berkeley-as-they-deliver-13895867.php

However, the report also quoted Sasha Iatsenia, the head of Kiwi Campus products, as explaining that human assistance was intentional and allowed robots to "observe" their surroundings without the need for expensive equipment such as lidar sensors.

Public information shows that the Kiwibots is made in China and assembled in the United States at a cost of only $2500 each.

The report pointed out that Kiwi Campus was angry at such a comparison similar to a "remote-controlled toy car", saying he was "parallel autonomy" and gave a statement:

The robot itself makes short-distance navigation between the road marks and uses artificial intelligence technology to avoid pedestrians and keep driving in the middle of the sidewalk.

With GPS, Colombian operators saw the robot on the street map, while airborne cameras showed them the robot's view of its surroundings, with a total delay of 150ms, Sasha Iatsenia said.

In other words, the online celebrity food delivery driverless car company is not completely self-driving, someone is behind it, and Kiwi Campus admits it.

So, the blame leaves only one question: have they publicly stated that the car is not self-driving, but manually operated?

Human-computer interaction and propaganda doubt

To answer the question, look for clues on Kiwi's website, and on the "Technology" page, they put a line like this:

We're mastering human-robot interaction.

We are improving human-computer interaction.

There is no mention that the unmanned car is self-driving, but in terms of technical interpretation, there is no indication of manual control behind it.

In May 2018, after interviewing Kiwi founder Felipe Chavez, DigitalJournal published an article explaining how the kiwibot robot works:

The KiwiBot uses binocular computer vision to navigate around the campus. However, it is also aided by humans who take control if the robot needs help crossing the street or navigating through busy areas.

Kiwibot uses computer vision to navigate around the campus. If a robot needs help crossing the road or crossing a busy area, humans will also help control it.

Reference link:

Http://www.digitaljournal.com/tech-and-science/technology/kiwi-robots-deliver-food-for-hungry-berkeley-students/article/523271#ixzz5y37oDRxo

This statement means that people will only intervene to help when the robot is in trouble. It is different from the "manual road sign machine to complete short-distance navigation" as pointed out in the San Francisco Chronicle.

As for whether you should wear the label of "impersonating AI", maybe you have your own answer more or less.

No artificial, no intelligence.

However, the doubt encountered by Kiwibot may also be one of the concerns about the landing of AI at present.

Since the fame of AlphaGo World War I, AI has become the hottest word and the darling of media, capital and technological innovation.

So under the upsurge, it is inevitable that fish and dragons are mixed, and bubbles are inevitable.

Since the beginning of this year, many companies that have raised funds under the guise of AI have been skinned.

But on the other hand, more and more people are beginning to realize the stage of technology landing. For a new technology with a sense of the future, such as "artificial intelligence", more people may have the imagination of "science fiction" and always feel that it can be achieved in one step.

For example, self-driving, is a fully intelligent car, where you want to go, omniscience, fully autonomous driving.

According to the trend of technological development, new technology always goes through a process of "overvalued in the short term and undervalued in the long term".

Kiwibot was kind of lying at gunpoint.

However, in fact, even if the autopilot lands, it will go through a "limited" to "unrestricted" landing process.

And in the current landing process, remote operation, also known as cloud driving, is inevitable.

The so-called "cloud driving" is that when the unmanned car is driving on the actual road, it is monitored by the remote control center and takes over the vehicle to complete the driving task when necessary.

This is also the consensus of regulators. When California opened applications for driverless cars in 2018, it was clearly stated that vehicles without safety guards should be equipped with remote control.

And this trend has spawned cloud-driven startups, such as Scotty Labs, which focus on developing remote control technologies for vehicles and have secured investment from Google.

At present, all the unmanned car companies that have launched the RoboTaxi grand plan have also set up remote operation centers one after another.

So for the progress of AI, sometimes "artificial + intelligence" is a joke, sometimes it is also the truth.

In fact, in 2019, the wind is like this, the landing is king, how much artificial, how much intelligence is no longer important.

What problems can be solved, how many problems can be solved, and what problems can be solved. Landing and value creation is the best way to test AI.

Perhaps after a while, it will no longer matter whether there is a "AI" tag and whether it is "artificial intelligence".

True AI, pseudo-AI, creating value more efficiently is a good AI.

What do you say?

-end-

Https://www.toutiao.com/a6736367779412181512/

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.