The ChatGPT wrapper product boom is an uncanny valley hellscape

We live in AI gold rush. Many companies tries to monetize trends. Standard behavior.<p>We have lived through some trends. Cloud computing, web30, metaverse, nft. Not all were bad, not all were transformative.<p>Since chatbots will kill google search as we know it, this time it is for real. Everyone will try to get slice of that pie.<p>One more thing is that some companies provide pages so vague, that you literally cannot understand what stand behind buzzwords. &quot;We enhance future with AI&quot;, &quot;we transform your business with AI&quot;. Corporate mumbo-jumbo. These page descriptions are written for the people who are tech illiterate. Effectively these sites are para-sites ;-)

I was thinking about this and my take is following; if somebody can provide me a browser extension that can help me while I &quot;surf&quot; the web, I would pay some tiny subscription fee of like $2.99 or $3.99 for it.<p>My vision of such browser extension was; let&#x27;s say I stumble upon some random website and I want to know what this website is all about, ChatGPT can summarize a website for me. I already tried prompting ChatGPT to summarize me some random websites and it gave me pretty good responses. Or I stumble upon some long article but I&#x27;m in hurry, I would want such article to be summarized for me, again ChatGPT can do it. Or I looked up and found some long podcast or interview on YouTube, this now can be summarized by Google&#x27;s Gemini better than ChatGPT.<p>I was thinking about making such extension but for me personally it would be a lot of work for not so much financially viable idea but I would like to have such extension and pay for it.

I wonder if it would be possible to create a &#x27;Dead Internet Google&#x27; by instructing ChatGPT to create a fake &#x27;search results&#x27; page for any given search page, and generating a fake website for each clicked linked.

I agree with the spirit of this article but it would’ve been a lot stronger with examples.<p>Generally speaking though, “prompt engineering” does not add sufficient value to merit a product in and of itself. No prompt is going to make a killer app. ChatGPT already is the killer app.

The celtic knot and cyan app icon boom.

Sparkle icon is devastated

The technical diagram of these shit companies really nailed it. 10&#x2F;10.

This post is complaining about something that has nothing to do with ChatGPT. They complain because in order to get your business to get indexed and found, you need a large amount of useless content. It is the same reason that every recipe you find online has 2 pages of drivel about their family and traveling somewhere before you get to the recipe. Blame SEO and the search indexers. That is why the market exists for this nonsense; it wasn&#x27;t created by LLMs.

First, none of the products differentiate themselves<p>Second, it&#x27;s hard to find good AGPL&#x27;d projects with communities. Most projects are building an open core enshittware model around MIT or Apache.<p>I think someone is going to need to write and open source AI Manifesto that speaks out against this low novelty open-washing strategy.

Based on the headline, I expected more of an exposition into the wrapper product landscape and accompanying boom than one oversimplification of one specific type of GPT wrapper.