From the beginning, Inaba spent heavily on research and development without concern for dividends—a corporate mission he described as “walking the narrow path.” But within three years, he and his team of 500 employees were shipping Fujitsu’s first numerical-control machine to Makino Milling Machine Co. In 1972, Fujitsu-Fanuc Ltd.—the “Fanuc” an acronym for Fuji Automatic Numerical Control—was founded as a separate entity, with Inaba in charge. […] The result of Nishikawa’s insight was the Fanuc Intelligent Edge Link and Drive, or Field. The system, introduced in 2016, is an open, cloud-based platform that allows Fanuc to collect global manufacturing data in real time on a previously unimaginable scale and funnel it to self-teaching robots. According to Fanuc, Field has already yielded advancements for tasks such as robotic bin-picking. Previously, the selection of a single part from a bin full of similar parts arranged in random orientations required skilled programmers to “teach” the robots how to perform the task. Now, Fanuc’s robots are teaching themselves. “After 1,000 attempts, the robot has a success rate of 60%,” a company release said. “After 5,000 attempts it can already pick up 90% of all parts—without a single line of program code having to be written.” Fanuc has so far declined to discuss its strategy concerning its venture into AI and machine learning. An employee who would only identify himself as Mr. Tanaka, because he wasn’t authorized to speak on the record, says the company will continue to focus on China. But, he adds, “we cannot rely on our past. As a company, we must adapt to new technology before we can create new technology. This will take time, but it’s necessary—the next generation of products have more functions, more connectivity, and more intelligence.”
wo summers ago, Courtenay Cotton led a workshop on machine learning that I attended with a New York–based group called the Women and Surveillance Initiative. It was a welcome introduction to the subject and a rare opportunity to cut through the hype to understand both the value of machine learning and the complications of this field of research. In our recent interview, Cotton, who now works as lead data scientist at n-Join, once again offered her clear thinking on machine learning and where it is headed.
Earlier this year our organization, the Rockefeller Family Fund (RFF), announced that it would divest its holdings in fossil fuel companies. We mean to do this gradually, but in a public statement we singled out ExxonMobil for immediate divestment because of its “morally reprehensible conduct.”1 For over a quarter-century the company tried to deceive policymakers and the public about the realities of climate change, protecting its profits at the cost of immense damage to life on this planet. Our criticism carries a certain historical irony. John D. Rockefeller founded Standard Oil, and ExxonMobil is Standard Oil’s largest direct descendant. In a sense we were turning against the company where most of the Rockefeller family’s wealth was created. (Other members of the Rockefeller family have been trying to get ExxonMobil to change its behavior for over a decade.) Approached by some reporters for comment, an ExxonMobil spokesman replied, “It’s not surprising that they’re divesting from the company since they’re already funding a conspiracy against us.”2 What we had funded was an investigative journalism project.
The only reason that anyone could be induced to take part in such a dangerous business was the fabulous profit that could be made. Gideon Allen & Sons, a whaling syndicate based in New Bedford, Massachusetts, made returns of 60% a year during much of the 19th century by financing whaling voyages—perhaps the best performance of any firm in American history. It was the most successful of a very successful bunch. Overall returns in the whaling business in New Bedford between 1817 and 1892 averaged 14% a year—an impressive record by any standard. New Bedford was not the only whaling port in America; nor was America the only whaling nation. Yet according to a study published in 1859, of the 900-odd active whaling ships around the world in 1850, 700 were American, and 70% of those came from New Bedford.
If ZunZuneo looks ridiculous in retrospect, it’s because 2011 is a different country. We now know U.S. security apparatus may threaten the “open Internet” as much as an oppressive government, if not more. Clinton’s speeches as secretary of state dwell on freedom of expression but not freedom from surveillance, and now—following the NSA revelations—we have a good idea why. Beyond all this, as sociologist Zeynep Tufecki writes, it’s likely that the failure of ZunZuneo will threaten online activism abroad, even if it’s not associated with the U.S. government.
And yet nobody wanted to add Peenemünde, where the Germans developed the V-2 rocket during the 1940s, to the glorious list of creative hothouses that includes Periclean Athens, Renaissance Florence, Belle Époque Paris and latter-day Austin, Texas. How much easier to tell us, one more time, how jazz bands work, how someone came up with the idea for the Slinky, or what shade of paint, when applied to the walls of your office, is most conducive to originality
We’ve treated ’scale’ like an unalloyed good for so long that it seems peculiar to question it. There are plenty of reasons for wanting to scale businesses and services up to make more things for more people in more areas; perhaps the strongest is that things usually get cheaper and quicker to provide. The problem is that scale has a cost, and that’s being unable to respond to the wants and needs of unique individuals. Theoretically, that’s not a problem in a free market, but of course, we don’t have a free market, and we certainly don’t have a free market when it comes to politics and media.