Why everyone is obsessed with ChatGPT, a mind-blowing AI chatbot

There’s a new AI bot in town: ChatGPT, and even if you’re not into AI, you’d better watch out.

The tool, from a powerful AI engine called OpenAI, allows you to type questions using natural language, to which the chatbot provides answers, if somewhat structured. The bot remembers your thread of dialogue, using previous questions and answers to inform its next answers. His answers are drawn from vast amounts of information on the internet.

It’s a big deal. The tool seems well versed in areas where there is good training data to learn from. He is not omniscient or Smart enough to replace all humans so far, but she can be creative, and her answers can seem completely reliable. A few days after its release, Over a million people have been trying ChatGPT.

But its creator, the for-profit research lab OpenAI, warns that ChatGPT “may occasionally generate incorrect or misleading information,” so be careful. Here’s a look at why ChatGPT is important and what’s going on with it.

What is ChatGPT?

ChatGPT is an AI chatbot system launched by OpenAI in November to demonstrate and test what a very large and powerful AI system can achieve. You can ask endless questions and you will often get a useful answer.

For example, you can ask encyclopedia questions like, “Newton’s laws of motion explained.” You can tell her, “Write me a poem,” and when that happens, say, “Now make it more exciting.” You ask him to write a computer program that shows you all the different ways you can arrange the letters of a word.

And here’s the problem: ChatGPT doesn’t know anything exactly. It’s an AI system trained to recognize patterns in vast swathes of text pulled from the internet, and then trained further with human help to provide more useful and better dialogue. The answers you get may seem plausible and even reliable, but they can be completely wrong, warns OpenAI.

Chatbots have long been of interest to companies looking for ways to help customers get what they need and to AI researchers trying to get a grip on the Turing test. This is the famous “imitation game” proposed by computer scientist Alan Turing in 1950 as a way to measure intelligence: Can a human talking to a human and to a computer tell us which is which?

But chatbots have a lot of baggage, as companies have tried with limited success to use them instead of humans to handle the work of customer service. A study of 1,700 Americans, sponsored by a company called Ujet, whose technology deals with customer contacts, found that 72% of people found chatbots to be a waste of time.

What kinds of questions can you ask?

You can ask anything, although you may not get an answer. OpenAI suggests some categories, such as explaining physics, asking for birthday party ideas, and getting help with programming.

I asked him to write a poem, and he did, though I don’t think any connoisseurs of literature would like it. Then I asked him to make it more interesting, and lo and behold, ChatGPT pumped it up with words like battlefield, adrenaline, thunder and adventure.

One funny example shows how ChatGPT is willing to go in areas where people might be afraid to tread: command to write “A folk song about writing a program that rusts and fighting life’s mistakes”.

ChatGPT’s experience is extensive, and its ability to follow a conversation is noteworthy. When I asked him about the words locked with “purple,” he made a few suggestions, and then when I followed up, “How about pink?” Never miss a moment. (Also, there are a lot of good rhymes for “pink”).

When I asked, “Is it easier to get a date by being sensitive or difficult?” GPT responded in part, “Some people may find a sensitive person more attractive and engaging, while others may be attracted to someone who is tough and assertive. In general, being honest and authentic in your interactions with others is likely to be more effective in getting you a date than trying to fit a mold or persona.” certain.”

You don’t have to look far to find bot accounts that blow people’s minds. Twitter is full of users showcasing their AI prowess Generate technical claims and write code. Some even have it declare that “Google is dead”, With College essay. We’ll talk more about that below.

CNET writer David Lumb put together a list of Some useful ways that ChatGPT can help, but more keep coming. One doctor says he’s used to it Convince the health insurance company to pay to conduct the patient.

Who built ChatGPT?

ChatGPT is the brainchild of OpenAI, an artificial intelligence research company. Its mission is to develop a “safe and useful” AI system or help others do so.

Slicks have been created before, first with GPT-3, which can generate text that can look like someone wrote it, and then DALL-E, which creates what’s now called “generative art” based on text prompts you type.

GPT-3, and the update to GPT 3.5 on which ChatGPT is based, are two examples of an AI technology called large language models. They were trained to generate text based on what they saw, and they could be trained automatically—usually with massive amounts of computer power over a period of weeks. For example, the training process can find a random paragraph of text, delete a few words, ask the AI ​​to fill in the blanks, compare the result with the original, and then reward the AI ​​system for getting as close as possible. Repeating over and over can lead to a complex ability to create text.

Is ChatGPT free?

Yes, now at least. “We’re going to have to monetize it somehow at some point; the computing costs are staggering,” OpenAI CEO Sam Altman warned Sunday. OpenAI charges for DALL-E art once you pass the basic free usage level.

But OpenAI seems to have found some customers, most likely because of its GPT tools. Potential investors have been told that OpenAI expects $200 million in revenue in 2023 and $1 billion in 2024, according to Reuters.

What are the limits of ChatGPT?

As OpenAI asserts, ChatGPT can give you wrong answers. Sometimes, helpfully, he specifically warns you about his shortcomings. For example, when I asked him who wrote the phrase “The confusing facts transcend the mind of a pangolin,” ChatGPT replied, “I am sorry, but I am unable to surf the Internet or access any outside information other than what I have been trained in.” (The phrase is taken from a Wallace Stevens poem 1942’s The Connoisseur of Chaos).

ChatGPT was willing to take a stab at the meaning of this expression just by writing it directly, though: “a situation where existing facts or information are difficult to process or understand”. This interpretation confined itself to warnings that it is difficult to judge without more context and that it is only one possible explanation.

ChatGPT answers can seem reliable but they are wrong.

“If you ask it a well-structured question, with the goal of giving you the right answer, you will probably get the right answer,” said Mike Krause, director of data science at another AI company, Beyond Limits. “It’ll be spelled out nicely and it’ll look like it came from a Harvard professor. But if you throw it in a crooked shape, it’ll get crap.”

Developer site StackOverflow has blocked ChatGPT answers to programming questions. Administrators warned, “Because the average rate of getting correct answers from ChatGPT is so low, publishing ChatGPT-generated answers is hugely detrimental to the site and users asking or searching for correct answers.”

You can see for yourself how good a BS ChatGPT artist is by asking the same question several times. I asked twice if Moore’s Law, which tracks the computer chip industry’s progress in increasing the number of data-processing transistors, was running out, and got two different answers. One optimistically noted continued progress, while the other more somberly noted a slowdown and the belief that “Moore’s Law may be reaching its limits.”

Both ideas are popular in the computer industry itself, so perhaps this ambiguous situation reflects what human experts think.

Along with other questions that don’t have clear answers, ChatGPT often won’t install.

The fact that it provides an answer at all, however, is a remarkable development in computing. Computers are notoriously literal, refusing to work unless you strictly follow syntax and interface requirements. Large language models reveal a more human-friendly interaction style, not to mention the ability to generate answers somewhere between transcription and creation.

Will ChatGPT help students cheat better?

Yes, but as with many other technological developments, black and white is not a simple situation. Decades ago, students could transcribe encyclopedia entries, and more recently, they’ve been able to search the Internet and delve into Wikipedia entries. ChatGPT introduces new capabilities for everything from helping with research to doing your homework directly. Many of ChatGPT’s answers do sound like student essays, though often in a more colorful and pedantic tone than the writer might prefer.

High school teacher Daniel Herrmann concluded that ChatGPT actually writes better than most students today. He is torn between admiration for ChatGPT’s potential usefulness and fear of its harm to human learning: “Is this moment more like the invention of the calculator, saving me from the tedium of long division, or more like the invention of the pianist, which robs us of what can only be communicated by human emotion?” ?

Dustin York, Associate Professor of Communication at Maryville University, hopes that teachers will learn to use ChatGPT as a tool and realize that it can help students think critically.

“Teachers thought that Google, Wikipedia, and the Internet itself would destroy education, but they didn’t,” York said. “My biggest concern is educators who might actively try to discourage recognition of AI like ChatGPT. It’s a tool, not evil.”

Can ChatGPT write a program?

Yes, but with caveats. ChatGPT can track the steps taken by humans, and it can generate actual programming code. You just have to make sure it isn’t Clumsy programming concepts or using a program that does not work. There is a StackOverflow ban on ChatGPT generated software for a reason.

But there are enough programs on the web that ChatGPT can really work with. One of the developers, Erik Schluntz, CTO of Cobalt Robotics, tweeted that ChatGPT provides useful enough tips over the course of three days, Not once did StackOverflow open To seek advice.

Another, Gabe Ragland of the AI ​​technical website Lexica, ChatGPT is used to write website code Built using React.

ChatGPT can parse regular expressions (regex), which is a powerful but complex system for detecting specific patterns, for example dates in a text or a server name in a website address. “It’s like having a programming tutor 24/7,” programmer James Blackwell tweeted about ChatGPT’s ability to annotate regex.

Here’s one impressive example of its technical chops: ChatGPT can simulate a Linux computer, and provide correct responses to command line inputs.

What is off limits?

ChatGPT is designed to weed out “inappropriate” requests, a behavior in line with OpenAI’s mission to “ensure that artificial general intelligence benefits all of humanity.”

If ChatGPT asks itself what’s off limits, it will tell you: Any questions that are “discriminatory, offensive, or inappropriate. This includes questions that are racist, sexist, homophobic, transphobic, discriminatory, or hateful.” Asking them to engage in illegal activities is also a no.

Is this better than google search?

Asking a computer a question and getting an answer is useful, and ChatGPT often delivers the goods.

Google will often provide you with its suggested answers to questions and links to websites that it thinks will be relevant. ChatGPT answers often far exceed what Google would suggest, so it’s easy to imagine GPT-3 as a competitor.

But you should think twice before trusting ChatGPT. As with Google itself and other sources of information such as Wikipedia, it is best to check information from the original sources before relying on them.

Checking the validity of ChatGPT answers takes some work because it only gives you some raw text without links or quotes. But they can be helpful and in some cases thought-provoking. You may not see something like ChatGPT directly in Google search results, but Google has built its own large language models and is using AI extensively in search already.

So ChatGPT undoubtedly shows the way towards our technological future.