State of the Blog

ChatGPT came out in November 2022 and it has completely changed the Internet. In my post Learning about ChatGPT via ChatGPT I had prompted: "Will ChatGPT replace Google?" to which the model back then outputted:

No, it is unlikely that ChatGPT will replace Google or other search engines in the near future. While ChatGPT is an advanced language model that can generate human-like responses to a wide range of questions and prompts, it is not designed to replace search engines.

Search engines like Google are optimized for searching large collections of text data and returning relevant results quickly, while ChatGPT is designed to generate text and answer questions based on a given context. While ChatGPT can be used to generate answers to specific questions, it is not capable of handling the scale and complexity of a global search engine like Google.

In addition, search engines like Google use a variety of algorithms and techniques, such as web crawling, indexing, and ranking, to return the most relevant results, while ChatGPT is trained on text data and generates responses based on patterns in that data.

Therefore, while ChatGPT is a powerful tool for natural language processing and text generation, it is not a replacement for search engines like Google.

Well, I don't have the numbers but I believe ChatGPT has completely replaced search engines, which is dominated by Google, for a lot of people. I do have numbers for my own site and my web traffic from actual people has fallen off a cliff, as I have put it before. The majority of traffic to this site are just bots: bots that have indicated that they are bots via their User-Agent header or bots that are masquerading as real users. For many years my site has been popular because I essentially handcrafted practical and useful bioinformatics tutorials but these days you can just ask ChatGPT or another tool to generate a more comprehensive tutorial in seconds.

I wrote a blog post on How and where to get information? after hearing that Stack Overflow (SO) and OpenAI have recently announced a partnership back in 2024 May. Like many other who code, I had relied heavily on SO and I believed back then that by making this deal with OpenAI, SO had essentially signed its own death note and I wondered where I would look for information.

Apparently you can run queries on SO and construct queries like how many questions are being asked each month. I downloaded the data from that query and for throwback reasons plotted the data using (base) R running in the terminal using my own brain and fingers to type the commands that I have learned many years ago:

R version 4.5.2 (2025-10-31) -- "[Not] Part in a Rumble"
Copyright (C) 2025 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

  Natural language support but running in an English locale

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> dat <- read.csv("QueryResults.csv", header=TRUE)
> plot(as.Date(dat$Created), dat$X, type="l", xlab='Date', ylab="Questions per month", main="StackOverflow questions per month")
> q()
Save workspace image? [y/n/c]: n

Now looking at the data, I guess SO already saw the decline and just decided to give up and sell its data in 2024 May. My own website's traffic has exactly the same trend as that plot, starting in 2010 but peaking around 2017/2018 and declining to the level of traffic of when I first started the blog. Unlucky for me, I didn't get an offer from OpenAI to sell them my blog's data (because they could just scrap it for free).

I believe that many websites, not just tutorial/documentation based ones, are experiencing this massive drop in traffic. I guess most people are aware of the paper "AI models collapse when trained on recursively generated data"; if less and less useful websites, i.e., data, gets generated because there's no longer an incentive to write high quality material, we will just keep spiralling downwards.

Recently, "Artificial intelligence tools expand scientists’ impact but contract science’s focus" was published in Nature. I'm not sure why the authors decided to publish this behind a paywall but at least you can read another version of the paper on arXiv. There is also a commentary in Science on the paper. As summarised in the pre-print:

On average, the use of AI helps individual scientists publish 3.02 times more papers, receive 4.84
times more citations, and become team leaders 1.37 years earlier. This substantial academic benefit may be a driving force behind the accelerated rate of AI adoption.

However, we also find unintended consequences from the increased prevalence of AI-augmented research. In all fields, AI-augmented research focuses on a narrower scope of scientific topics and reduces the scientific engagement of follow-on research, leading to more overlapping research works that slows the expansion of knowledge.

Further, with a greater concentration of collective attention to the same AI papers, the adoption of AI appears to induce authors to engage in collective hill-climbing, catalyzing solutions to known problems rather than creating new ones.

Personally, I didn't think the AI hype would last this long. I still feel like a lot of money is being thrown in a raging fire. I don't know how sustainable things are right now: scraping the entire Internet to produce centralised tools that provide "average" output that is polluting the Internet with said average content that is being fed back into model training.

Recently, I read that the main thing that drives economics is incentives. I still have an incentive to write because I like writing as it makes me think. Despite everything that is going on, I'll still keep blogging, even if I'm just writing mostly for myself.