Feb 15, 2019

Axios Future

Have your friends signed up?

Any stories we should be chasing? Hit reply to this email or message me at steve@axios.com. Kaveh Waddell is at kaveh@axios.com and Erica Pandey at erica@axios.com.

Okay, let's start with ...

1 big thing: Deepfakes go bigger

Illustration: Rebecca Zisser/Axios

Researchers have broadened the controversial technology called "deepfakes" — AI-generated media that experts fear could roil coming elections by convincingly depicting people saying or doing things they never did.

Kaveh reports: A new computer program, created at OpenAI, the San Francisco AI lab, is the latest front in deepfakes, producing remarkably human-sounding prose that opens the prospect of fake news circulated at industrial scale.

  • As we have previously reported, existing technology can already create fake video, audio, and images — AI forgeries that together could transform the misinformation industry.

Details: The new OpenAI program is like a supercharged autocomplete — a system like Google uses to guess the next words in your search, except for whole blocks of text. Write the first sentence of a sci-fi story and the computer does the rest. Begin a news article and the computer completes it.

  • The quality varies. Sometimes, the text is a bit confused — but when it gets it right, the results are stunning. It has enormous potential for fiction or screenwriting. (See bonus item below.)
  • But it could also be used for dystopian ends — huge, coordinated onslaughts of racist invective, fake news stories and made-up Amazon reviews.
  • "There are probably amazingly imaginative malicious things you could do with this technology," says Kristian Hammond, a Northwestern professor and CEO of AI company Narrative Science.

Why OpenAI created the program: The lab says it aims to create a general model for language — a program that can do what humans can (speak, understand, summarize, answer questions and translate) without being specifically trained to do each.

  • In terms of what can go wrong, this is ultimately a dual-use question — AI researchers are working on this technology for one reason, but bad players can use it for another.
  • "I think you have to ask yourself what systems like this might be capable of in several years," says Jack Clark, OpenAI's policy director. The quality of the program's writing is likely to keep improving as more data and computing power is added to the mix, he says.

How it works: The program "writes" by choosing the best next word based on both the human-written prompt and an enormous database of text it has read on the internet.

  • One reason the program is so convincing is that it was trained with enormous amounts of computing power and data — resources out of reach of many less wealthy research organizations.
  • An important point: The AI writer can only make stuff up. It can't tell the difference between a fact and a lie, which is part of what makes it volatile. Figuring out how to "teach" it what's true remains a huge challenge, says Hammond.

Because of how this technology could be abused, OpenAI announced that it will not release the program that generates the most convincing-sounding text. (Next week, we will return to this aspect of the story.)

The big question: How dangerous are text deepfakes?

  • AI-generated images have reached a level where they're often indistinguishable from real photographs. That's not true for generated text, which can sometimes be incoherent.
  • One new door that AI-generated text opens: "Conversational," rather than "broadcast" deepfakes, says Aviv Ovadya, a misinformation researcher who founded the non-profit Thoughtful Technology Project. It could be used to influence people one-on-one on a massive scale, rather than distributing a small number of forgeries widely.
Bonus: AI wrote this story

Photo illustration: Rebecca Zisser/Axios, Photo: Jim Watson/Getty

Kaveh wrote the first two sentences in italics below — they're from a story we published in Monday's newsletter. Then, the OpenAI program discussed above wrote the rest, on the first try.

  • We added bolding and bullet points to mimic Axios style, but didn't change the computer-generated text at all.
  • A very important note: Everything you'll read below, after the italicized intro — quotes, "facts" and all — is not true. The AI program invented them.

On the heels of a sweeping new U.S. plan to retain dominance in artificial intelligence, the Pentagon has cast Chinese development of intelligent weapons as an existential threat to the international order.

A day after the release of an executive order by President Trump that omits naming China, the Defense Department, in a new AI strategy document, speaks in stark terms of a "destabilizing" Chinese threat.

It warns of a "new arms race in AI" and says the United States "will not sit idly by" as a "highly advanced new generation of weapons capable of waging asymmetric warfare" is "possessed by aggressive actors."

Related: New White House plan on China may spark a cyber arms race

"China uses new and innovative methods to enable its advanced military technology to proliferate around the world, particularly to countries with which we have strategic partnerships," the Pentagon said in its five-page strategy outline last week.

The new U.S. strategy will be a major component of the White House's first National Security Strategy, coming in two parts in September.

  • The first part, which Trump wants to roll out in full by June, is a much more traditional geopolitical playbook focused on the threats posed by China and Russia.
  • The second part of the NSS is "a broad review of all U.S. national security interests, and the potential new relationships needed for those interests," the document said.

"The President has directed me to undertake a study of our strategy toward a world of artificial intelligence," Defense Secretary James Mattis told the Senate Armed Services Committee on Thursday.

2. Alipay in America

Illustration: Sarah Grillo/Axios

Alibaba, China's e-commerce behemoth, is making new inroads in the U.S. and Canada, adding its mobile payment system to thousands of stores — convenience, drug stores and other shops — frequented by Americans of all income levels.

  • Erica writes: Alipay insists that it is not after American customers, instead seeking to serve Chinese tourists, students and business people who come in droves to North America every year.
  • But while its partnerships — with chains like 7-11 and Walgreens — may serve a lot of Chinese visitors, these stores' main clientele is ordinary locals, who now will become acquainted with the Alipay name.

The big picture: Alibaba is enticing U.S. businesses with Alipay, which allows users to pay by scanning a QR code on the app. It has more than 700 million Chinese users. It is telling U.S. companies that if they add the system, they'll get a big piece of the tens of billions dollars Chinese nationals spend in the U.S. every year.

Alipay already has 4 million U.S. users and works with luxury brands like Lacoste and Rebecca Minkoff, and can be used at duty-free shops and Vegas hotels — places frequented by Chinese tourists.

But the move this week into 7,000 Walgreens, and earlier into 7-Eleven in Canada, is part of a separate strategy, says Humphrey Ho, an analyst with the Hylink Group. "The real play here is for the students, the new immigrants, and the people here to work for just a few years," Ho says.

  • There's a massive market of Chinese nationals who are in the U.S. longer than tourists, but would still like to use Alipay instead of applying for a credit card.

My thought bubble: Though it's possible that Alibaba could follow Japan's strategy from the 1960s and 1970s and break open the American market, it seems to me that Alipay's expansion will be limited to Chinese nationals for the foreseeable future.

  • As a Walgreens shopper, I'm not sure just seeing that Alipay is an option would make me want to use it. Americans love their credit cards way too much.
  • Plus, it's unlikely that Alibaba will ever get approval from the U.S. to take on the role of a bank and dole out mobile wallets. The privacy concerns are too big, Ho says.
3. What you may have missed

Photo: Underwood Archives/Getty

Long week? We've got you covered. Here are the top stories from Future.

  1. Trump's AI strategy: It's all hands on deck.
  2. A new era for Amazon: Unfamiliar — and hostile.
  3. The anti-Amazon: Walmart's new role in society.
  4. What New Yorkers think about HQ2: Whiplash in Long Island City.
4. Worthy of your time

Illustration: Rebecca Zisser/Axios

The weak link in the global economy (Fergal O'Brien, William Horobin — Bloomberg)

Hospitals reap windfalls from outpatient drugs (Bob Herman — Axios)

India floats Chinese-style censorship (Vindu Goel — NYT)

Investors are scurrying back to Rome (Kate Allen — FT)

Why baths incubate ideas (The Economist)

5. 1 cartoon thing: Watching Peppa, talking like a Brit

Screenshot: Peppa Pig

In a new phenomenon, young Americans are falling under the influence of a cartoon pig, and speaking with British accents and idioms.

The culprit is Peppa Pig, the star of an animated U.K. series whose dialogue includes phrases not used by ordinary Americans, like knackered, snuggle, straightaway and toh-MAH'-toh, writes Kelly Conaboy at the Cut.

Go deeper: Watch Peppa Pig.