The A.I.War - How I Learned to Stop Worrying and Love the Bomb

CTO Lounge
1

IMAGE ALT TEXT HERE

Google Panics Over ChatGPT [The AI Wars Have Begun]

By now, many of us will be familiar with ChatGPT, and some of us may even get a laugh at some of the silly claims people in the media may be making about ChatGPT. What is not silly is the tangible threat to Google’s market dominance in Internet Search.

First off, I generally recommend ColdFusion videos. They are high quality, relevant, and interesting, and this one comes at an important time in the ChatGPT Media Bandwagon Circus. For Google, the threat comes from various sources, but a significant threat comes from Microsoft, and their intent to integrate the OpenAI technology into their Bing search engine. This could seriously undermine Alphabet’s (Google’s parent company) profit model on internet search advertising.

While much of the media exploit sensationalism created by imagined Artificial Intelligence opportunities, exploits, and threats, it is still too early to really understand all these imaginings. But, when Google calls a Code Red company meeting to discuss the threat the Microsoft OpenAI alliance poses, we can be sure Google are not just playing Chicken Little.

DOJ files second antitrust suit against Google, seeks to break up its ad business

This is not a good time for Google as they are also facing serious Department of Justice Antitrust allegations, similar to DOJ allegations against Microsoft that almost lead to the breakup of the company years ago.

image

It will be interesting to see how Alphabet/Google navigates these emerging threats to its market dominance.

Personal Perspective

Personally I have used Google Search for decades, almost since its inception, and I still use it today as my main search engine. But that does not mean I like it.

  1. I am looking for an effective replacement. Over the years the value of Google Search has definitely been decreasing for me. Increasingly the search results lead me to low value, and often useless content.

  2. In playing with ChatGPT I have found the results often more satisfying than Google Search, higher value content, with better context. And currently, no advertising! :partying_face:

    • I look forward to playing with the new Bing Search, to see how well it works
    • I look forward to seeing Google Search enhanced with similar ‘A.I.’ technology

  3. I generally look forward to such ‘paradigm shifts’ and I am sure that many people other than Google and Microsoft will be applying Generative Predictive Transformers all over the place.

  4. Transformation does not apply to just text, but also graphic arts, music, etc.

    • Already we are seeing people make claims about writing software, writing code, via GPT technology.

    • However, personally, I am curious how we could apply GPT technology to data such as JSON, XML, YAML, etc. :thinking: DISCLAIMER: I am expressing a personal curiousity, and not any insight into Forgerock Research & Development.

Not a Panacea

In playing with ChatGPT I recognize various weaknesses.

For example, one day I asked it “What is the meaning of life, the universe, and everything?” I got back a great answer, but I did not get the answer I wanted “42.” In the feedback to ChatGPT I noted that not only did it miss a great cultural/literary reference, but it has no sense of humour.

  • How well would effective handling of culture, literary, and other implied references work in GPT?

    • How can we train for this?

  • How in the world do we model humour in machine learning? :face_with_monocle:

    • Can this even be trained?

Others have noticed that ChatGPT has distinct biases, especially with respect to political content.

  • Arguably, there is bias in almost everything, but how do we minimize it or try to be fair?

  • Will we see progressive and conservative versions of ChatGPT and friends, the way we see progressive and conservative media?

    • What effect will the automated ‘echo chambers’ have on society?

    • I myself wrote a science fiction novel where the protagonist creates and markets a web browser plugin called the “Truth Machine” that filters all web pages through an A.I. that flags contents as true, false, and maybe. I stole that name from the novel The Truth Machine.

Microsoft, Google, and others have noted that when generating text and images, they will have to ensure that explicit adult content is not expressed.

  • This could be especially embarrassing to explain to a Kindergarten class. :scream:

  • But what else might be inappropriate generations?

Caveates

We have yet to see how A.I. such as ChatGPT can be used for evil, but some conjecture and speculation seems indicated.

  • If Google can monetize free Internet Search via advertising, via targeted advertising, how will they monetize A.I. capabilities in augmented search?

    • When monetizing A.I. speech, can the US First Amendment be used to protect Google and others from litigation?

  • Decades ago, since the time of Eliza, psychologists and others observed that many people are more willing to share personal and intimate information with a computer than with a real person.

    • As we have seen with Facebook and others, collecting, selling, and otherwise exploiting personal information has been very profitable, and very problematic.

    • As people become comfortable having discussions with search engines such as Bing and Google, instead of only making search queries, what might we reveal about ourselves that can be exploited for profit?

    • What happens when ChatBots start asking questions in return?

As we look forward to the new improved Bing, and Google’s response, it can be useful to consider both the utopian and dystopian consequences of introducing new technology.

Important Fact: Before the invention of the automobile, there were no deaths from automobile accidents.

2 Likes