ChatGPT Becomes One: How OpenAI's chatbot changed technology forever

There have been a few before and after moments in the age of modern technology. It was all one way, and then, it suddenly became clear that it would never be that way again. Netscape showed the world the Internet; Facebook has made this internet personal; The iPhone demonstrated how the mobile era would take over. There are other things — there's a dating app moment somewhere, and Netflix starting to stream movies might also qualify — but not much.

ChatGPT, which OpenAI launched a year ago, was perhaps the least game-changing of all. No one got on stage and announced that they had invented the future, and no one thought they had launched the thing that would make them rich. If we've learned one thing over the past 12 months, it's that no one — not OpenAI's competitors, not general technology users, not even the platform's creators — thought ChatGPT would become the fastest-growing consumer technology in history. In retrospect, the fact that no one saw ChatGPT coming may be exactly why everything seemed to change.

In the year since ChatGPT was launched, it has brought change to practically every corner of the tech industry. In a year marked by a significant decline in venture capital investment, any company with “AI” in its pitch seems able to raise money — $17.9 billion in just the third quarter of this year, according to Pitchbook, and some startups. The largest venture capital firms in the industry are raising huge funds just to continue pouring money into AI.

There appear to be a few companies already at the head of the pack: Anthropic is shaping up to be one of OpenAI's best and most well-funded competitors, Midjourney's image-generating AI is improving at a remarkable pace, and even just this week Pika He appeared out of nowhere Using a very cool AI video tool. But whether you love note-taking apps, audio mixing tools, or easy ways to summarize meetings, books, or legal documents, there's something new and great being released practically every day.

Meanwhile, on the other side of the technology industry, AI has consumed the largest companies on the planet. Microsoft, an OpenAI partner and investor, is betting big on AI-powered Bing while also bringing AI “assistants” to Office, Windows, Azure, and more. Google, which invented much of the core technology that is now suddenly ubiquitous, was quick to launch Bard and its generative search experience, integrating Duet AI into its workplace products. AI has been the focus of Amazon's announcements this year, from Alexa powered by LLM to one million new AI tools for AWS customers. The Meta now sees AI as an important part of its future, perhaps more so than the Metaverse. Nvidia's AI hardware has made it one of the most valuable companies on Earth. Even Apple, which has moved less aggressively among the tech giants, is starting to talk more about its AI efforts, and may have big plans for Siri soon. I could go on. Call it a boom, call it a bubble, but it's been a long time since the entire tech world was obsessed with one thing.

However, make no mistake: ChatGPT is the big winner in the ChatGPT revolution. It doesn't sound like much — its new audio and video features are great, but it's still mostly just a roughly designed chat interface — and it's been plagued by reliability issues, but that hasn't stopped its momentum. It had 1 million users in five days, 100 million after just two months, and now boasts 100 million every week.

ChatGPT and its model quickly became a billion-dollar business for OpenAI. It has achieved something almost impossible: it is a data provider, it makes money from other companies that want to build things on top of GPT models, and it is a successful consumer application in its own right. People pay $20 a month to use ChatGPT, while other companies pay a lot more to use their models – OpenAI makes them come and go.

You can't practically click a link on the internet anymore without encountering confident predictions about how AI will change everything. He can write your emails for you! It will sweep the Internet with the crap it generates! He can write code! He will write malware that destroys everything! He can make Pixar movies! You will be stuck in the uncanny valley forever! You will never have a job again! You will never……….. ever Need Post again! Artificial intelligence will save us! Artificial intelligence will kill us!

It's worth pausing here just for a second to point out that most of this technology is, in fact, still not very good. Big language models “hallucinate”, which is a nice way of saying that they make things up all the time. If you look at an AI image for more than about two seconds, you can always tell it was generated. The emails he writes to you are almost always automated in nature. AI systems are not smarter than humans, or more creative, or anything else. Is it remarkable that they are as good as they are? certainly! But so far AI is shaping up like self-driving cars: it's getting good faster than anyone thought, and it's going to be a hell of a work until it gets good enough to be everywhere. There is absolutely no reason, at the moment, to believe that we will reach some kind of superhuman artificial general intelligence any time soon. If ever.

This is the point where “no one saw this coming” gets complicated. AI may not be dead yet, but it's already better than most people expected. Even in recent weeks, OpenAI has split in two due to the speed at which ChatGPT has grown and OpenAI has moved to monetize it through the App Store and other tools. CEO Sam Altman was briefly forced to resign, for reasons we still don't know exactly: Was it a power play between board members and executives, the result of a disagreement over safety, or something else entirely?

This drama was bizarre, high-stakes, and perhaps ultimately unrelated to the broader question. So let's get back to this broader question: What are we actually building here?? Because all of this happened so quickly, and because the impacts of AI are so wide-ranging, 2023 will force everyone to catch up on what it all means.

OpenAI's original mission statement was to “advance digital intelligence in a way that is most likely to benefit humanity as a whole, without being constrained by the need to generate financial return.” It's mysterious, but it sounds good! It is also easy to determine when there is no financial return, and is more difficult when analysts estimate that the total addressable market is more than $1 trillion.

Across technology, and around the world, many people are thinking about this same tension. If you're Google CEO Sundar Pichai, and you've now spent five years unironically saying that AI is “deeper than electricity or fire,” do you have a responsibility to maximize its value to shareholders or to humanity? There's not a lot of evidence to show you can do both at once, and historically, shareholders tend to win. If AI is going to change everything, like literally everything, can it do so within the technology industry and economy as we know it? Is the AI ​​we need the same AI that makes the most money?

Sure, we'd love to be able to write work emails more quickly, and we'd love to be able to ask Excel to “turn this into a bar graph” instead of digging through menus. We like to be able to program by simply telling ChatGPT what we want our application to do. But do we want SEO-optimized and AI-generated news stories to replace the posts we used to love? Do we want AI robots that act like real characters and become anthropomorphic companions in our lives? Should we think of AI as a tool or as a collaborator? If an AI tool can be trained to create the song/movie/image/story I want right now, is that art or is it dystopia? Even as we begin to answer these questions, AI technology always seems to remain one step ahead and a cultural revolution ahead.

At the same time, there have been lawsuits accusing AI companies of plagiarizing artists' works, and several US judges have basically said that our current copyright laws don't know what to do with AI at all. Lawmakers have puzzled over the safety of AI, and President Joe Biden signed a fairly broad executive order directing agencies to set safety standards and companies to do good, not evil. There is an argument that the AI ​​revolution was built on unethical and/or illegal grounds, yet the creators of these models and companies continue to confidently move forward with their plans, while saying that it is impossible and anti-progressive to stop or slow them down.

I know this all gets really fast. The reality is that no one knows where all of this will be even 12 months from now, especially the people making the highest predictions. All you have to do is look at recent hype cycles — blockchain, the Metaverse, and many others — for evidence that things don't usually go the way we think. But there's so much momentum behind the AI ​​revolution, and so many companies deeply invested in its future, that it's hard to imagine GPTs going the way of NFTs.

If anything, the next 12 months of the AI ​​industry will move faster than the last 12 months. OpenAI technology has improved significantly since ChatGPT's first release, as have its competitors. The entire industry has had a year to think about all the places AI might be useful in our lives and profitable in their products. There will be new companies building AI chips, AI data centers, and the rest of the massive infrastructure required to make LLM work at speed and scale. We'll be getting a slew of AI-focused tools, like the Humane AI Pin, as companies try to figure out if chatbots can propel us into the post-smartphone era. (Although I personally won't be betting against monitors any time soon.)

We don't know yet whether artificial intelligence will ultimately change the world the way the Internet, social media, and smartphones have. These things were not just technological leaps, they actually reorganized our lives in fundamental and irreversible ways. If the ultimate form of AI was “my computer writes me some of my emails,” then AI would not be on that list. But there are a lot of smart people out there betting trillions of dollars that this is the beginning of the AI ​​story, not the end. If they're right, the day OpenAI launches a “research preview” of ChatGPT will be much more than a product launch for the ages. It would be the day the world changed, and we didn't even expect it to happen.