Today we want to talk about using AI tools to clone or copy existing websites.
We will start by providing an overview of artificial intelligence: what AI is, what is the current state of AI technologies and what are the AI limitations that you need to be mindful of.
Then we are going to discuss AI website builders, the advantages of using them instead of regular website builders, and the pros and cons of cloning existing websites with these tools.
Finally, we will explain why we believe that you don’t need a website in order to make money online, why building one is probably a waste of time if you are just starting out, and what you should do instead!
What You Need to Know About Artificial Intelligence (AI)
Let’s start by going over the most important things that you need to know about artificial intelligence.
We recommend reading this section in its entirety so that you have a solid understanding of the current state of AI technologies and their limitations.
However, if you aren’t interested in this subject and want to get straight into AI website builders, you can simply skip to the next section.
What is Artificial Intelligence?
The term “artificial intelligence” – “AI” for short – refers to a machine’s ability to perform cognitive tasks such as learning from experience, solving novel problems, and making decisions.
If you want to learn more about this subject, we recommend going through Google’s “AI Essentials” course that covers the current state of artificial intelligence technologies, their most common use cases in the business context, and their limitations that you need to be aware of.
This course also covers prompt engineering – the best practices for creating prompts when working with large language models such as Google’s Gemini.
You can watch the first lecture on YouTube:
What’s the Difference Between General Artificial Intelligence and Specialized Artificial Intelligence?
It’s important to understand the distinction between general artificial intelligence and specialized artificial intelligence:
- The term “general artificial intelligence” refers to machine intelligence that either matches or surpasses human intelligence. You may be familiar with this concept from TV shows like “Black Mirror” and movies like “The Matrix”, “Ex Machine” and “Her”. At this point in time, general artificial intelligence only exists in science fiction, though some computer scientists who specialize in AI argue that it might become a reality in the foreseeable future.
- The term “specialized artificial intelligence” refers to machine intelligence that can perform well in a specific cognitive skill or across a range of cognitive skills. Chess apps would be an example of the former, while large language models like ChatGPT would be an example of the latter. Software like this can outperform humans within its domain, but it cannot match or surpass human intelligence outside of it.
Software companies that use specialized artificial intelligence in their products are capitalizing on the fact that the general public associates the term “AI” with general artificial intelligence.
This is especially true for companies that sell access to chatbots powered by large language models.
For example, OpenAI, the company that created ChatGPT, has faced a massive backlash after seemingly intentionally using a voice similar to that of Scarlett Johansson’s to voice their chatbot without her consent.
Some people speculate that the reasoning behind this move was to create an association between ChatGPT and the sentient chatbot in the hit movie “Her” which was voiced by Johansson.
That’s an especially egregious example but it can be argued that the current hype around AI is partially fuelled by marketing that hints at general artificial intelligence, even if the companies aren’t making such claims outright.
If you want to successfully use AI tools in your business, you need to stay grounded in reality. These apps may be impressive but they haven’t caught up with humans in terms of intelligence yet. And they are almost certainly not sentient.
So don’t fall for all this science fiction marketing hype!
What are Large Language Models?
We have already mentioned large language models several times. But what are they?
Large language models (LLMs) are a type of generative AI – artificial intelligence whose primary purpose is to generate content – that works with language and can generate text in response to user queries.
This is the technology behind cutting-edge AI software such as OpenAI’s ChatGPT, Anthropic’s Claude, and Google’s Gemini.
Andrej Karpathy is a computer scientist who specializes in AI research. He was one of the founding members of OpenAI.
If you want to learn more about LLMs, his one-hour lecture about the subject is probably the best place to start:
What are the Limitations of Large Language Models?
Here’s a quick summary of how large language models function at this point in time:
Data goes in -> Something happens within the large language model that we cannot see -> Data comes out.
In other words, no one really knows how LLMs work, not even the people who created them. These AI systems are mysterious black boxes!
For example, as AI Safety advocate Robert Miles put it, we basically have no idea how ChatGPT does what it does. The same goes for all other LLMs out there.
Consequently, due to the inscrutability of these AI systems, the computer scientists who created them cannot ensure that their output is accurate.
This problem is known as “hallucinations”. Sometimes, the output of LLMs is correct, but other times, it’s either factually incorrect or just entirely made up.
There’s an entire AI research subfield called interpretability that is dedicated to understanding how these artificial intelligence systems work.
For example, Anthropic is doing interpretability research on their own LLM, Claude, and publishing papers about it!
However, while we have seen some progress in interpretability in recent years, it appears to be highly unlikely that computer scientists will figure out how large language models work anytime soon.
In fact, some AI skeptics argue that hallucinations are an inherent property of large language models and therefore cannot be fixed. According to them, hallucinations are a feature, not a bug!
Examples of Large Language Model Use Gone Wrong
Here are some examples of large language model use gone wrong:
- Countless students around the world have turned in nonsensical essays written by ChatGPT.
- A lawyer in New York who used ChatGPT while representing a client got in trouble with the judge because ChatGPT made up a bunch of court cases that did not exist.
- A supermarket chain in New Zealand introduced an LLM-powered meal planner app which then started providing deadly recipes to its users, such as a recipe for chlorine gas that it called an “aromatic water mix”.
- Chevy dealership’s ChatGPT-powered chatbot sold a car for $1 to one of its users and stated that it was a legally binding agreement. .
- Google’s experimental “AI Overview” search feature became a laughing stock when it started providing ridiculous responses to search queries, including advising its users to eat “at least one small rock per day” and explaining that although you cannot use gasoline to make spaghetti cook faster, you can use it to “make a spicy spaghetti dish”.
And let’s not forget that time when ChatGPT started spouting pure gibberish at its users for no apparent reason!
How Should You Use Large Language Models in Your Business?
Hallucinations make large language models completely unreliable: you cannot trust their output, especially when it comes to important tasks that can have serious consequences.
Practically speaking, this means that if you want to use large language models in the business context – or any other context, for that matter – you must ensure that their output is always fact-checked by a human with relevant domain expertise.
You can do it yourself if you have the requisite knowledge or you can outsource it to experts. What’s important is that there’s a human in the process that has the ability to spot hallucinations.
The best way to take advantage of LLM-powered AI tools is to use them to generate the first draft of whatever it is that you need: a website, a blog article, a sales page copy, etc.
Once you have that first draft, someone needs to edit it, fact-check it, and proofread it before it goes live.
This approach can help you drastically speed things up, especially if you take the time to master prompt engineering so that you can provide clear instructions to LLMs.
Never EVER Let Large Language Models Interact With Your Customers!
If you want to use a chatbot to reduce the customer support workload, make sure to use a regular one, not an LLM-powered AI one.
Remember, LLMs are black boxes, their behavior is unpredictable and there’s no way to guarantee that they will provide accurate information to your customers.
Moreover, they might end up giving not just bad but outright dangerous advice, saying offensive stuff or just rambling incoherently. Would you hire someone this unstable for a customer support role? Probably not.
Finally, there are quite a few people out there who love “jailbreaking” LLMs – getting these AI systems to behave in ways that they shouldn’t – and posting screenshots on Twitter. It’s a hobby for them.
If these jailbreaking enthusiasts catch wind that your chatbot is powered by an LLM, they are going to have a field day with it. Your company might end up going viral for all the wrong reasons as a result, like that Chevy dealership did when their chatbot sold a car for $1.
Also, something to keep in mind here is that the general public only has a vague understanding of what “AI” is, so if your ChatGPT-powered chatbot does something outrageous, your customers are not going to blame OpenAI for it, they are going to blame you.
In short, at this point in time, using LLM-powered chatbots in your business is just asking for trouble, so don’t do it!
What You Need to Know About AI Website Builders
What are AI Website Builders?
AI website builders are apps that use artificial intelligence to generate websites. It’s probably safe to say that most of them are powered by popular LLMs such as ChatGPT.
What are the Advantages of Using an AI Website Builder Instead of a Regular Website Builder?
While you don’t need to know how to code in order to use regular website builders, you still have to make various design decisions in order to create the first draft of your website.
In contrast, with AI website builders, all you need to do is provide a prompt describing what kind of website you want and the software will generate the first draft of that website for you.
The latter approach is a much faster way to create websites so the main advantage of using AI website builders is speed of development!
How Should You Use AI Website Builders in Your Business?
You should use them the same way you would use any LLM-powered software:
Have it generate the first draft of your website for you, then use the website editor to customize it so it would match your brand.
We also recommend hiring a web developer to manually inspect the code of your website just to make sure that it doesn’t contain hallucinations that might cause problems later down the road.
How to Clone an Existing Website With AI Website Builders
AI website builders that have a cloning functionality typically allow you to simply copy-paste the URL of the website that you want to clone and then generate a copy of that website for you that you can then customize.
The Pros and Cons of Using AI Website Builders to Clone Existing Websites
The main pro of using an AI website builder to clone existing websites is the speed of development.
If you see a website design that you like, you don’t need to waste time recreating it yourself, you can simply copy-paste the URL and the AI website builder will do it for you.
However, this approach to building websites also has several cons that you need to take into consideration:
- Ethical concerns – On the one hand, there’s that saying “Good artists copy, great artists steal”. On the other hand, outright plagiarizing someone else’s work is unethical.
- Reputational risks – If it comes out that you cloned a competitor’s website, you might get negative media coverage and have to deal with the ensuing public backlash. That might damage your brand.
- Legal risks – You might get sued for copyright infringement if you clone a website whose design has been copyrighted (eg. if you clone a copyrighted WordPress theme).
Also, it’s worth noting that when you are copying a website, you don’t have the context required to understand the reasoning behind its design.
You also don’t have access to the relevant business data so you can’t know how well that website is performing. It might look good but does it actually convert?
So you might end up copying a competitor’s design that won’t work for you or that isn’t even working for them to begin with.
Keep in mind that imitating your competitors without understanding why they are doing what they are doing is a terrible idea. Why would you want to follow their lead instead of walking your own path?
Ultimately, we wouldn’t recommend cloning other people’s websites unless you intend to keep the same overall structure but change everything else to the point that the design becomes unrecognizable.
But then wouldn’t it make more sense to simply use AI to generate the website that you want from scratch?
How to Choose the Right AI Website Builder for Your Business
There are so many AI website builders out there. So how can you choose one that’s right for your business?
All AI website builders offer functionality that allows you to generate websites using artificial intelligence.
You want to look at what other features a given AI website builder offers, whether your business could benefit from those features and how would they fit into your overall business workflow.
Speaking of your business workflow, you also want to look for integrations that would allow you to easily integrate a given AI website builder with various other apps that you are already using. Native integrations are ideal but at the very least there should be a Zapier integration.
Pricing is another important factor but you shouldn’t simply compare the price tags of different apps and go with the cheapest option. Instead, ask yourself how would the cost of a given AI website builder fit into your overall software budget considering the features that this tool would provide.
User interface probably doesn’t matter that much if all you need is one website but if you intend to be building websites on a regular basis, then it’s something that you should pay attention to.
In that case, it makes sense to sign up for free trials of several AI website builders so that you can play around with them and see which user interface you like the most.
Also, before you make the final decision, you might want to check out customer reviews on websites like TrustRadius, Capterra and G2. That can reveal potential problems that you wouldn’t be able to spot during a free trial, such as subpar customer support.
Hot Take: You Don’t Need a Website to Launch an Online Business!
We would argue that if you are just starting out, building a website is a waste of resources. Why?
Because you don’t even know if there’s a demand for your product or service!
Here’s what we advise doing instead:
- Create an offer
- Build a sales funnel for that offer
- Start driving traffic to that sales funnel with paid ads
This will enable you to quickly get market feedback, validate your business idea, and start generating revenue.
Realistically, you will probably find that your initial offer doesn’t quite resonate with your target audience.
In all likelihood, you will probably need to test several different offers until you find one that has a product-market fit.
Fortunately, ClickFunnels 2.0 includes an AI sales funnel builder that allows you to generate the first draft of your sales funnel in just 10-15 minutes.
Once you have that first draft, you can customize it with our funnel editor.
Then, once your sales funnel is live, you can use our A/B testing functionality to optimize it for conversions.
All this makes it easy to continue experimenting with various business ideas until you find a winning one.
Once you are sure that your dream customers want what you are selling, then you can create a website for your business with our website builder!
You can generate the first draft of your sales funnel in just 10-15 minutes with our AI sales funnel builder!
Try ClickFunnels 2.0 Risk-Free!
Let’s be real:
We are biased in favor of our software.
So we understand if you take what we say about it with a grain of salt.
Fortunately, we have a free 14-day trial, so you don’t have to take our word for it.
You can go and see ClickFunnels 2.0 for yourself without any risk!
So what are you waiting for? 🧐
