Skip to navigationSkip to content
Close
Trump has a plan to keep America first in artificial intelligence

Trump has a plan to keep America first in artificial intelligence

Read more on MIT Technology Review

Contributions

  • DARPA has shown that a government and private enterprise partnership model for innovation can achieve great things.

    If done properly, the one thing that is a precursor to AI work is to make sure that national data estates are available in efficient structures. There are loads of data sitting in tape

    DARPA has shown that a government and private enterprise partnership model for innovation can achieve great things.

    If done properly, the one thing that is a precursor to AI work is to make sure that national data estates are available in efficient structures. There are loads of data sitting in tape archives in basements around many government agencies. Again, an opportunity for public-private partnership to bring all of this data into broadly available data lakes that can be consumed by those looking to advance their AI/ML efforts.

  • It’s hard to keep “America first” in AI when all the major breakthroughs are coming from companies that operate independently of Washington.

    Not a problem that Chinese leadership has.

  • I just can’t be emphatic enough that this arms race around AI will have disastrous consequences if it is not designed to a major degree with individual control over the data that will power it. Otherwise we will all be locked into increasingly mega platforms controlled by companies or governments (or

    I just can’t be emphatic enough that this arms race around AI will have disastrous consequences if it is not designed to a major degree with individual control over the data that will power it. Otherwise we will all be locked into increasingly mega platforms controlled by companies or governments (or both at the same time for much of the world) with less and less choice and understanding of how we are being manipulated. I truly believe our free will and liberty is at stake.

  • It's great to see this kind of commitment and recognition to the fact that AI is here to stay and will continue to shape and transform our existence. Is this a counter to China's investment in AI? - absolutely! The clue is in the wording "consistent with American values" which BTW is fine as long as

    It's great to see this kind of commitment and recognition to the fact that AI is here to stay and will continue to shape and transform our existence. Is this a counter to China's investment in AI? - absolutely! The clue is in the wording "consistent with American values" which BTW is fine as long as those values align with ethical and fair treatment through automated use of AI.

  • The US is many days late and a few billion dollars short—China has a commanding lead in AI research and funding. The US does have a private sector champion in Google, but this kind of PR window dressing isn't going to to move the needle.

  • They need lots and lots of money, and an innovative spirit.

    In present day dollars, the Apollo project cost $22 billion.

    Manhattan project cost $144 billion.

    Like all the new frontier projects before it, artificial intelligence has to be well funded, and Congress has to be on board.

  • I think this is good. The way I see it, the biggest focus is to get the MOST PEOPLE researching and thinking. China and the US are both spurred by startup capital, but a governmental focus by China is likely to get a lot more heads together. We need this too. (By number of research papers published

    I think this is good. The way I see it, the biggest focus is to get the MOST PEOPLE researching and thinking. China and the US are both spurred by startup capital, but a governmental focus by China is likely to get a lot more heads together. We need this too. (By number of research papers published, China has come way up from behind, but Europe as a whole is big too — see https://qz.com/1490424/europe-publishes-more-ai-papers-than-the-us-or-china/.) edit to add: https://qz.com/work/1547302/the-best-cities-and-countries-to-live-in-if-you-want-a-job-in-ai/

  • Excellent first step. Very aspirational outline, lots of doors open for individual agencies to embrace AI initiatives.

    We will certainly need and want more. But you have to start somewhere. And it’s better to be non-specific when beginning on a journey to a non-specific destination. This is very different

    Excellent first step. Very aspirational outline, lots of doors open for individual agencies to embrace AI initiatives.

    We will certainly need and want more. But you have to start somewhere. And it’s better to be non-specific when beginning on a journey to a non-specific destination. This is very different than putting a man on the moon. The path to AI leadership will unfold over time.

    I think we may end up looking back at this time as a turning point. Not due to this order, but technology ultimately becoming an enabler and uniter. Technology shifts the balance of power, and very well may tilt that balance away from the silos of governments towards the open fields of the masses.

  • This is quite a turn from the Trump administration—but it may be the president's itch for competition and one-upping China that's driving this AI push.

  • President Trump will always put the US 1st.

  • Seeing the words “Trump” and “Intelligence” is so jarring.

  • Ya, good luck persuading others...

    'Engage internationally: It will call for a strategy for international collaboration that ensures AI is developed in a way consistent with American “values and interests."'

  • It's an arms race plain and simple. Lots of past scientific advancement has been rooted in military applications. The space race is a great example of this. Should the government be involved in the development of super intelligent AI? I think that questions pales in comparison to the ones we will be

    It's an arms race plain and simple. Lots of past scientific advancement has been rooted in military applications. The space race is a great example of this. Should the government be involved in the development of super intelligent AI? I think that questions pales in comparison to the ones we will be asking if some other state actor accomplishes that goal first. Someone is going to do it eventually, do we want to risk being on the losing end of whatever that leads to?

  • Like everything he does, he no doubt saw a Fox News segment on China's AI investment and hurriedly threw this half-baked idea together.

  • Great ideas and use of funding. But I would like more specifics on exactly where the money is going towards and how this will benefit the electorate.

  • The whole AI community is talking how to combat biases in AI and now we have an “American AI”!!!

    .... Just kidding!

  • I always try and situate the present within the tapestry of our connected history. Before commenting on this specific initiative, and proclaiming to be a success or failure, let’s look at other broad presidential initiatives and see how they succeed or failed and why. We can probably put the Manhattan

    I always try and situate the present within the tapestry of our connected history. Before commenting on this specific initiative, and proclaiming to be a success or failure, let’s look at other broad presidential initiatives and see how they succeed or failed and why. We can probably put the Manhattan Project and Man on the Moon as successes. Beyond the obvious of, “it was a lot of money,” what other mechanisms, organizational configurations, and leadership made it succeed?

    How about some more recent presidential initiatives, such as, to keep our “automotive” and “supercomputer” leads? Were they successes? If so, was it because they poured federal dollars into the initiative, they convened and lead a coordinated initiative? Was it the coordination among private companies, government agencies and/or a combination.

    Now, with those analytics lenses, let’s try and assess how successful the plan to keep America first in AI will be. However, there may need to be another pre-question. How do we determine the many dimensions of AI, is it the hardware, ie new CPU’s GPU, software, ie unsupervised neural nets, is it quantum computing, etc?

  • Legitimate question:

    Should the state be involved with AI development and implementation?

  • He don't need no intelligence!