The internet used to belong to cats.
It was the late 1990s, a weird time when everyone online was posting pictures of cats (believe it or not, my cat hates you).(opens in new tab)The blog that pioneered posting pictures of grumpy cats is still online). start a blog. Post a picture of your cute cat companion. It would be nice if someone left a comment.
Of course, there were things online other than cats. But while the internet, spelled with a capital ‘I’, was big at the time, it was still not essential and no one was entirely sure what it would be like.
I was there first. Somewhere in 1994, my father connected his Windows 3.1 PC to the Internet via his US Robotics dial-up modem and said,
You can move files (very slowly) via FTP, search documents via Gopher, and the best browser was called Cello. I used it to find some strange poems, early William Gibson works, Tool song lyrics. I knew the World Wide Web would one day be huge, but at the time it was still empty. The Internet in those days was not yet for everyone. It was a vast frontier filled with many things to explore.
After the first era of exploration, the Internet was no longer one thing, it was all things. Soon it became difficult to complete school projects or work without the Internet. Businesses, institutions and homes are all online. Facebook and Twitter are here. Some people find themselves spending a lot of time on the Internet. Some continued to post pictures of their cats.
I am writing this very short history of the Internet because I feel strongly that we are once again at the end of an era.Despite all the good and bad things it has brought us, the internet has always been mostly oursYes, some of the emails you receive are automated, Google bots scour the web for information on how to rank web pages in search results, and some computers mine bitcoin. turned into a botnet. But much of the Internet was designed to be used and explored by humans, or at least to feel like it.
Get away, human. AI is here to take over
OpenAI recently released a new version of its AI model that powers its ChatGPT chatbot, GPT-4. For the first time, it has the ability to understand and process information from images. This opens up some previously unavailable possibilities, and we’ve already seen examples of how everything changes.
Developed by OpenAI, which grew from a non-profit lab to a commercial artificial intelligence powerhouse, ChatGPT is an AI language model that can generate surprisingly human-like answers to a large number of queries. Chatbots have been around for decades, but ChatGPT was different. You can answer both specific and vague questions alike. You can impersonate someone else. You can write poetry and plays in the style of famous poets and playwrights (see here for a more comprehensive overview of ChatGPT).
The previous major version of ChatGPT, based on the GPT-3 model, metaphorically destroyed the internet, quickly becoming the hottest tech topic and fastest growing app of all time.
GPT-4, on the other hand, has the potential to literally destroy the Internet. We won’t know that soon. When chatting, the new ChatGPT, based on his GPT-4 model, doesn’t sound radically different from its predecessor. But it paves the way for much more powerful applications.
AI expert Travis Fischer gives some examples of what AI can do. thread(opens in new tab)These include building AI that can interact with all or most. web page elements(opens in new tab)Not only can you create a functional website by simply taking input from a sketch of real world notebook(opens in new tab)The changes in GPT-4 allow it to understand longer inputs and produce longer outputs. This means, for example, that you can write your own fairly complex apps.
Tweet may have been deleted
(opens in new tab)
(opens in new tab)
Unlike humans, who can do one or perhaps a few tasks at a time, AI bots are limited by the constraints of the hardware on which they run (and associated costs, which we’ll omit for simplicity). . With a powerful processor and lots of memory, he could do the same thing millions and billions of times a day if he wanted to. When considering the impact of a tool such as GPT-4, one must consider the enormous scale at which it can operate.
Currently, ChatGPT (and soon similar AI bots) use web page and build People started using it to browse the web without webpages, then built the web without webpages, and finally built the web designed to be used without them. The first phase will start immediately. It’ll be a while before we get to the last one, but I’m sure it’s coming. At that point, the web will be a huge, impossibly tangled web of data, largely impermeable to humans unless accessed using AI assistants.
Need something online? Ask AI
For example, let’s say you need to browse a number of real estate listings in Houston. Perhaps you are considering a purchase or need real estate data for a work project.Good news: AdeptAILabs is built a bot(opens in new tab) can do it for you. In another example, the company’s AI bots can: Perform complex tasks in Salesforce(opens in new tab)which requires only one sentence as input.
Tweet may have been deleted
(opens in new tab)
(opens in new tab)
If you are studying for a law school exam, I recommend asking ChatGPT for help. After all, it beats most law graduates on the bar exam, according to the study.(opens in new tab)AI-based legal assistant(opens in new tab) Tools based on GPT-4 already exist, and while they cannot replace human lawyers, they may soon become essential tools for lawyers.
In fact, there are already AI assistants that can perform all sorts of web tasks on behalf of users, and some are about to do so. more powerful(opens in new tab) Thanks to GPT-4. Once they can perform certain tasks just like humans, doing things the old way without the help of AI bots can feel like a waste of time. And when that happens, perhaps the web will begin to change.
another kind of web
One direction this could all go is a kind of meta-web where we rarely actually browse the web. Instead, talk to a bot that hits the web to get what you need. Like fully self-driving cars that no longer need wheels, the web may soon begin to change and adapt to landscapes that no longer need to look good to the human eye. Instead, it should be easily accessible to bots. In a way, this is already happening. Much of what happens online is done by software, not by humans. It may be in another format, such as an adjusted summary.
An early (and somewhat disturbing) example of this event is found in Captcha, a troubling puzzle where you must prove you are human. Recently, I noticed a new kind of capture that uses AI-generated images. There is considerable irony in a system that prompts an AI to create images of a robot and then requires a human to accurately identify these images in the hope that another AI will not be able to do so. I have. In this equation, humans feel like unnecessary intermediaries that one day will be completely removed.
This has already happened. To test her GPT-4 skills, OpenAI had her ChatGPT pose as a blind person.(opens in new tab) Avoid captcha prompts nicely. In fact, a human employee of TaskRabbit convinced him to solve the captcha and provide an answer on his behalf. The AI passed tests that only humans were designed to pass, tricking humans into bidding.
This does not mean that ChatGPT is “alive” or smarter than us in any way. It’s just a tool that humans use. However, the ability of GPT-4 to perform certain tasks easily and quickly foreshadows a future in which it makes no sense for humans to perform such tasks at all. Although a little dystopian, it’s easy to imagine a captcha designed to scare away humans as well as stupid bots. The web designed for use by artificial intelligence may contain parts that were not designed for use by humans. Not without AI mediation.
However, there is still one thing you can do yourself. Post a picture of your cat.
