Concerns about the rise of A.I.!

TUESDAY, MAY 2, 2023

One milestone has long since been passed: Geoffrey Hinton is profiled on the front page of today's New York Times.

Hinton is identified by the Times as "The Godfather of A.I." At the start of today's report, Cade Metz describes the arc of Hinton's work—and notes Hinton's current state of high concern:

METZ (5/2/23): Geoffrey Hinton was an artificial intelligence pioneer. In 2012, Dr. Hinton and two of his graduate students at the University of Toronto created technology that became the intellectual foundation for the A.I. systems that the tech industry’s biggest companies believe is a key to their future.

On Monday, however, he officially joined a growing chorus of critics who say those companies are racing toward danger with their aggressive campaign to create products based on generative artificial intelligence, the technology that powers popular chatbots like ChatGPT.

Dr. Hinton said he has quit his job at Google, where he has worked for more than a decade and became one of the most respected voices in the field, so he can freely speak out about the risks of A.I. A part of him, he said, now regrets his life’s work.

What are the risks to which Hinton alludes? In his report, Metz attributes several such concerns to Dr. Hinton. 

At the risk of seeming cheeky, we'll focus on this concern:

METZ: Dr. Hinton’s journey from A.I. groundbreaker to doomsayer marks a remarkable moment for the technology industry at perhaps its most important inflection point in decades. Industry leaders believe the new A.I. systems could be as important as the introduction of the web browser in the early 1990s and could lead to breakthroughs in areas ranging from drug research to education.

But gnawing at many industry insiders is a fear that they are releasing something dangerous into the wild. Generative A.I. can already be a tool for misinformation. Soon, it could be a risk to jobs. Somewhere down the line, tech’s biggest worriers say, it could be a risk to humanity.

“It is hard to see how you can prevent the bad actors from using it for bad things,” Dr. Hinton said.

[...]

His immediate concern is that the internet will be flooded with false photos, videos and text, and the average person will “not be able to know what is true anymore.” 

The average person will “not be able to know what is true anymore?” 

At the risk of sounding cheeky, our flailing nation passed that checkpoint a long time ago! (And not because of applications of A.I.)  Nor is it clear that this problem is limited to "the average person."

Can we talk? Thanks to the so-called "democratization of media"—that is to say, thanks to the rise of talk radio, partisan cable, the Internet and social media—the average person has long since stopped "being able to know what's true anymore."

The termination of Tucker Carlson will barely begin to correct the drift in this direction. 

The spread of partisan misinformation and disinformation became a very big business decades ago. As we've been saying for many years, given the giant profits involved in this culture of mis- and disinformation, we don't see an obvious way out of this growing mess.

For what it's worth, we'll add this additional point:

Forget the much-maligned "average person!" It isn't clear that our journalistic and academic elites have the ability "to know what's true anymore," or that they ever did. 

For the latest relatively minor example, we would cite this recent column in the New York Times concerning the question cited in its title:

How Well Does Masking Work? And Other Pandemic Questions We Need to Answer.

This is the latest column discussing whether masking worked "at the population level" during the pandemic without explaining what the technical term "at the population level" actually means. 

By now, we think we know what that term means in the present context, but we can't say we're entirely sure. What we can tell you is this:

We've been amazed, for the past few months, by how many reports we've seen in which this topic is explored without the author explaining the meaning of that key term. This latest murky attempt at settling this important issue comes from a high academic figure at Brown who's been put into print by the Times.

(For what it's worth, Carlson seemed to delight in spreading misinformation or disinformation about the effectiveness of masking and masks. It seemed to be one of his favorite very strange games.)

Beyond all this lies the later Wittgenstein's suggestion that much of the western world's highest order "philosophical" writing has been based on types of "conceptual confusion." Add that suggestion to the stew and you have a sprawling, unfortunate backdrop to Dr. Hinton's immediate concern.

Reader, please! The average person won't "be able to know what's true anymore?” 

Through no fault of the average person, we passed that milestone long ago! It's true over there within the red tribe, but it's also true among us blues. Beyond that, our journalistic and academic elites are routinely in the dark too.

Could A.I. really make matters worse? If so, the godfather of A.I. really should be deeply concerned!

Concerning the later Wittgenstein: As always, we link you to Professor Horwich's brief overview for the New York Times, from way back in 2013:

Philosophy is respected, even exalted, for its promise to provide fundamental insights into the human condition and the ultimate character of the universe. It’s taken for granted that there is deep understanding to be obtained of the nature of consciousness, of how knowledge of the external world is possible, of whether our decisions can be truly free, of the structure of any just society, and so on...

If so, then we are duped and bound to be disappointed, says Wittgenstein. For these are mere pseudo-problems, the misbegotten products of linguistic illusion and muddled thinking.

Truer words never said?


23 comments:

  1. The second amendment is evil.

    ReplyDelete
  2. Says Somerby today: nobody knows anything and experts cannot be trusted.

    How does it help anyone in our society to believe this?

    Somerby offers two sources of proof for this statement: (1) Wittgenstein's complaints about language in philosophy, (2) a controversy during a pandemic involving a new virus in which science had to come up with guidelines before having investigated the threat.

    Do these two examples really tell us that nothing can be known for sure and there is no source of truth in our society? I don't think so.

    For one thing, have standards for differentiating truth and lies between individual people changed? For another, has the scientific method of empirically testing hypotheses changed? Is data no longer data? Has mathematics been abandoned? Have people lost the ability to differentiate between opinion and fact? Is reality any different than it was?

    We have always had the capacity to tell lies. We are confronted with AI which can quickly produce mountains of info without any regulation on what is true and what is false. But we have always had the ability to produce and spread lies. How have people protected themselves from lies in the past?

    1. We have established standards for trust and applied them in the world.
    2. We have developed means of testing info to see whether info is accurate. One of these is science.
    3. We have developed protective procedures for ensuring that what we encounter is not harmful (meat inspections, password access, vaccines, employee badges).
    4. We have informal systems for checking with others to share experience about important life tasks, products, events. Reviews are an example of this.
    5. We have laws to punish those who take advantage of others by lying. We also have police and investigators.
    6. People are inherently risk averse and will limit the chances they take under uncertain circumstances. People will become more skeptical and less willing to engage if sources are not trustworthy.

    System to protect people from untruth may lag the telling of lies, but we will develop means of protection as the need arises.

    Somerby's doom and gloom is the equivalent of luddite fears about other forms of technological progress. In spite of the lack of limitations on lying in our society today, more than half the populace has not bought into right wing disinformation and outright lying. I suspect the number still believing Trump after his trials will be a much smaller percentage. Eventually, we may enact laws to exclude con men from running for office. In the meantime, newspapers will do more research into local level candidates to prevent another George Santos. RFK Jr. will not gain traction.

    Liars are being tolerated in the arenas where there is profit to be made off them. Politicians may discover that lying hurts them in the long run, so there may ultimately be a natural check and balance in the form of a reactive public. Just as companies that provide defective products will lose their long-term customers.

    Somerby's constant claims that the media and experts cannot be trusted, should fall on deaf ears once people realize that it is only those with expertise who can discern the truth when unscrupulous people are telling mass lies. Meanwhile, Somerby is hurting everyday people when he tells his readers to distrust the more trustworthy elements of society. As people find out that Somerby (and his ilk) are unreliable, then they will stop paying attention to what he says. I believe that is already happening with most of his readers. The rest will suffer until they start paying attention to whether Somerby's complaints about the media hold water.

    You know what's real and true by testing info. If you test what Somerby says here, it flunks. If there are more lies being told, there will be more testing to be done, but people will figure it out.

    ReplyDelete
    Replies
    1. Thanks for a well thought-out comment, @3:03. You asked, "has the scientific method of empirically testing hypotheses changed? Is data no longer data? Has mathematics been abandoned? Have people lost the ability to differentiate between opinion and fact? Is reality any different than it was?"

      Sadly, I'm afraid the answer is sometimes "yes". One example is global warming. Serious sanctions are often applied to scientists who don't follow the preferred narrative.

      In specific situations, this is happening in other scientific fields. One critic compared it to Lysenkoism in USSR under Stalin. Lysenko drove scientists who disagreed with him out scientific leadership and entirely out of science. Because of this structure, bad science persisted for along time. Among tragic results was mass starvation, because agriculture research was based on invalid science.

      Another example is a de-facto ban on honest research on racial differences in intelligence.

      Delete
    2. The findings of researchers studying global warming are not "narratives," preferred or otherwise. Lysenko was incorrect in his findings and his research was not scientific. The major problem in Russia was its belief in pseudoscience and superstition. There were more reasons for starvation than bad science, including bad organization of a newly centralized land management plan, corruption, and a triumph of politics over other considerations.

      There has been ongoing research on intelligence. The ostracism of those using research to support racist views was never a ban on their research. Jensen was not banned or fired. He was argued with and protested but his work was published. It was also used by segregationists to justify their activities.

      Delete
    3. About those serious sanctions. Provide examples. You seem to be confusing the activity of science with that of politics in the state of Florida. I've flushed more well formed material down my bathroom toilet. It takes a rarefied form of ignorance to make up shit in order to justify opposing the consensus of roughly 98% of climatologists.And calling the results of their work a "narrative". A narrative, FYI , is a well scripted right wing message that has been manipulated to look like science to the uneducated, like the anti-vaccine narrative currently being promoted by the Florida Republicans.

      Delete
    4. you do know you will never get a serious response from DinC, right?

      Milton Friedman and Thomas Sowell told him something half a century ago, and nothing's going to change his mind now. He is locked in.

      Delete
    5. I'll always appreciate David in Cal making the point that police are running a protection racket on the citizens.

      Delete
  3. In Wittgenstein's time, people didn't read him. They went on holding conversations without the problems he warned about, and there was no falling apart of knowledge or society. Then Wittgenstein died. Meanwhile, modern philosophers have found the loopholes in Wittgenstein's work and advanced philosophy on a sounder foundation with respect to language, and all is fine and dandy in philosophyland.

    ReplyDelete
  4. Here is something real to worry about:

    "Republicans Push to Change Modern Divorce Laws
    May 2, 2023 at 12:59 pm EDT By Taegan Goddard

    Rolling Stone: “Republicans across the country are now reconsidering no-fault divorce. There isn’t a huge mystery behind the campaign: Like the crusades against abortion and contraception, making it more difficult to leave an unhappy marriage is about control.”

    ReplyDelete
    Replies
    1. Right wing darling Steven Crowder has been pushing a reconsideration of no fault, but then a video was released of him viciously verbally abusing his pregnant wife; he is now going through a divorce where fault abounds, at least on his actions, making the issue moot in his case.

      Delete
    2. The entire comment at Rolling Stone was "Rolling Stone: “Republicans across the country are now reconsidering no-fault divorce. There isn’t a huge mystery behind the campaign: Like the crusades against abortion and contraception, making it more difficult to leave an unhappy marriage is about control.”
      No examples or links are provided.

      Delete
    3. Link takes you to "The Next Front in the GOP's War on Women: No-Fault Divorce by Tessa Stuart, May 2, 2023. Lots of details there.

      Delete
  5. Republicans should support statehood for Puerto Rico and the District of Columbia.

    ReplyDelete
  6. Hinton’s role in AI is being a bit overblown. (The “godfather” moniker may in fact be self generated!) Hinton is at the end of his career, after selling out to Google a decade ago, with an eye towards his legacy.

    AI causes consternation among the ruling elites, who are mostly right wing, because a) they want the cheap labor AI might provide and yet b) they recognize that controlling humans via labor is what allows them to maintain their dominance.

    Somerby’s minders have told him to go out and try to manufacture fear and ignorance on the issue of AI.

    Fear mongering aside, what has AI wrought? So far we have a false image of the pope wearing a puffer jacket, yet no one has forgotten that in reality the pope does indeed wear a funny hat. The AI generated response to comments here was a failure, making obvious errors and was easily defeated. There’s zero credible evidence Jesus existed yet here we are thousands of years later and people remain enraptured by a myth.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. I agree with you about Jesus, @4:22. There's also zero credible evidence that liberal programs like rent control do more good than harm, yet people also remain enraptured to this myth.

      Delete
    3. Good for who?

      Delete
    4. Oh I see, so we're picking on Jesus now. How convenient since he isn't here to defend himself. And a fellow Jew at that. Where's the solidarity? Seems a brazenly anti-Christian statement to be making in a public space. Maybe tone it down a little or at least give equal time to dissing the prophet Mohammed if you've got the balls for that kind of thing. Lucky you don't live in Florida or DeSantis would seek you out and and put you on a bus. After cancelling your season Disney pass. So maybe instead of singling out the guy who atoned for your sins consider the broader picture here and whether there is any credible evidence to account for your own divine belief system, unless you are an atheist or agnostic, in which case congratulations.

      Delete
    5. Did they exist?

      0 < p(Abraham) < p(Jesus) < p(Mohammed) < 1

      Delete
    6. They did not.

      Also rent control works.

      https://thehill.com/opinion/finance/573841-theres-no-denying-the-data-rent-control-works/

      Delete
    7. Whenever there is a debatable issue in which data supporting both sides of an argument are available online or in print, such as the rent control issue, or whether raising minimum wage adversely affects the job market, or virtually anything having to do with the black population, DIC will invariably ignore the existence of data that does not support his claims, opting for the position espoused by right wing pundits, and going directly to a pronouncement of certainty with no acknowledgement of the complexity of the issue or the fact that the debate is ongoing and not settled. There is no fixing the short- circuited mind.

      Delete
  7. We live in a giant spiral galaxy.

    ReplyDelete