Ask HN: How will SEO break ChatGPT?

5 points by TorontoTheGood 3 years ago | 4 comments

How will SEO break ChatGPT? Are there other risks that may have us ending up with an increasingly unusable product just like standard web search has turned out?

sharemywin 3 years ago |

Why would you trust ChatGPt about facts in the first place?

And as for product recommendations it's really generic.

what is the best camera?

As an AI language model, I do not have personal preferences, opinions, or biases. However, I can provide you with some information on what makes a good camera.

The "best" camera for you depends on your specific needs and preferences. Different cameras are designed for different purposes, such as professional photography, vlogging, or travel photography. Some factors to consider when choosing a camera include:

Image quality: The camera's sensor size and resolution can affect the quality of the images.

Lens selection: The availability of interchangeable lenses can enhance the versatility of the camera.

Size and weight: Consider the camera's portability if you plan on carrying it around for extended periods.

User-friendly interface: The camera should have an intuitive interface that is easy to use.

Battery life: Consider how long the camera can last on a single charge.

Price: The price of a camera is an important factor to consider.

Some popular camera brands that are known for producing high-quality cameras include Canon, Nikon, Sony, Fujifilm, and Panasonic. It is important to do your research and read reviews to find the camera that best fits your needs and budget.

nigamanth 3 years ago |

Currently, GPT3 is trained on a lot of data, but all this data is pre-2021. Essentially, no matter how much "new content" comes out on the web, ChatGPT won't know about it.

Nothing can influence ChatGPT's dataset or training data for now. However, GPT4 will have more data points than GPT3, and by the time they train more LLMs on GPT4, the data on the web will seep into ChatGPT.

The same way SEO realizes on measures, the same way ChatGPT answers rely on transformer layers which choose the most succinct answers, given enough feeding of the wrong data and no fine-tuning, large LLMs could break.

zamnos 3 years ago | |

Is it simply a repetition count? If I made a billion pages that said the Earth was only 6,000 years old and stuffed that into the corpus, would that override the million pages saying it's 4.543 billion years old instead? Is there no PageRank-like algorithm that tracks what other links are saying about it?

bediger4000 3 years ago |

I think that since SEO and corporate blogging will use ChatGPT to generate content that they think will move sites up in the rankings, LLMs will end up being trained on rubbish. There will be a feedback loop, and later generations of LLMs will experience something analogous to "heterodyning".