Some users are accusing OpenAI of slowing down its AI-backed chatbot, ChatGPT. The American AI company has been making headlines for all the wrong reasons for a while now.
After firing its founder and CEO Sam Altman, the company rehired him in the same position after OpenAI employees threatened to quit.
Reportedly, OpenAI fired Altman to stop him from releasing a human-threatening AI model called Q*. While the OpenAI debacle is still settling down, ChatGPT users have started reporting that OpenAI is intentionally slowing down the service.
Some users have taken to Reddit and X (formerly Twitter) to report that the recently unveiled GPT4 Turbo has stopped responding to prompts like it used to before. The word on the street is that OpenAI is nerfing ChatGPT to stop people from misusing it.
ChatGPT users claim the service is becoming lazy
According to a considerable number of ChatGPT users, the AI-backed bot has stopped responding to questions in-depth as it used to.
In one of the examples, X user Eric Hartford asked the chatbot to use Kotlin and OpenAI APIs to develop a streaming client. Rather than writing the code, ChatGPT gave Hartford steps on how he could solve the problem.
Wow GPT4 has been seriously nerfed. I just tried interactively developing a bit of code, something that worked perfect 2 weeks ago, and it resisted and acted lazy. Ay, I'm going to have to turn to open models for coding... (maybe a good thing tho)https://t.co/KFSz2md4Dr
— Eric Hartford (@erhartford) November 28, 2023
Hartford pointed out that ChatGPT would have actually solved the problem a few weeks back by directly writing the code whenever asked. Moreover, when he specifically asked ChatGPT to write code, the bot provided incomplete sets of code.
Another X user, Rohit, said ChatGPT has become lazy and incompetent, further noting that the service has started taking shortcuts when it comes to completing lengthy tasks.
OpenAI has safety-ed GPT-4 sufficiently that its become lazy and incompetent.
— rohit (@krishnanrohit) November 28, 2023
Convert this file? Too long. Write a table? Here's the first three lines. Read this link? Sorry can't. Read this py file? Oops not allowed.
So frustrating.
The X user asked ChatGPT to write a table from some data but the AI bot only wrote the first 3 lines and asked him to complete the rest. Aside from this, the service refused to read certain files, which it used to earlier as per the post.
Likewise, several Reddit users are reporting that ChatGPT has started refusing to complete even simple tasks. For instance, Reddit user Acceptable-Amount-14 asked ChatGPT to fill a CSV file of 15 entries with 8 columns each but the AI bot complained that the data was too much to process.
Responding to an X user, OpenAI's API Product Manager Owen Campbell Moore noted that a bug was causing these issues. However, the company is working to fix this, he added.
GPT-4 has become so @#$%@ lazy, it won't even output more than a handful of code lines now:
— Benjamin De Kraker (@BenjaminDEKR) November 28, 2023
Fix. This. pic.twitter.com/RHBgEwIVnW
Adding some sarcasm to his reply, Moore admitted that ChatGPT asking him to write his own code is driving him crazy too. It is possible he was simply trying to relate to the ChatGPT users' frustration. Alternatively, Moore could be pointing out that users are pushing ChatGPT too much.
In line with this, Wikipedia's founder Jimmy Wales recently noted that ChatGPT is terrible when it comes to writing articles on the free online encyclopedia.
