uniladtech homepage
  • News
    • Tech News
    • AI
  • Gadgets
    • Apple
    • iPhone
  • Gaming
    • Playstation
    • Xbox
  • Science
    • News
    • Space
  • Streaming
    • Netflix
  • Vehicles
    • Car News
  • Social Media
    • WhatsApp
    • YouTube
  • Advertise
  • Terms
  • Privacy & Cookies
  • LADbible Group
  • LADbible
  • UNILAD
  • SPORTbible
  • GAMINGbible
  • Tyla
  • FOODbible
  • License Our Content
  • About Us & Contact
  • Jobs
  • Latest
  • Archive
  • Topics A-Z
  • Authors
Facebook
Instagram
X
TikTok
Snapchat
WhatsApp
Submit Your Content
ChatGPT breaks down days after Sam Altman warned people to stop using new feature

Home> News> AI

Published 09:55 3 Apr 2025 GMT+1

ChatGPT breaks down days after Sam Altman warned people to stop using new feature

'Our GPUs are melting'

Tom Chapman

Tom Chapman

google discoverFollow us on Google Discover
Featured Image Credit: Bloomberg / Contributor / Getty
AI
ChatGPT
Tech News

Advert

Advert

Advert

Well, we can't say Sam Altman didn't warn you. While creating all those cute Studio Ghibli-inspired avatars was fun for a while, it seems that karma has hit ChatGPT after the legendary Hayao Miyazaki previously branded the idea of using artificial intelligence to make art as 'disgusting'. We were told that ChatGPT's servers were struggling to keep up with demand, and apparently, it's led to a global outage.

As the race between Grok and ChatGPT continues to heat up (like the latter's GPUs), Sam Altman warned against using ChatGPT's new AI image generation feature. The rollout of ChatGPT's GPT-4o model on March 25 was originally praised as a marked improvement over what was previously on offer, and with ChatGPT landing 1,000,000 new users in just an hour, it looked like Altman was going to reap the rewards.

Altman previously warned that ChatGPT was struggling (Bloomberg / Contributor / Getty)
Altman previously warned that ChatGPT was struggling (Bloomberg / Contributor / Getty)

Sadly, ChatGPT got too big for its boots and soon went kaput. Wannabe Ghibli artists were greeted by a message saying "something went wrong" and were advised to refresh the page...to no avail.

Advert

Downdetector revealed a massive spike in reports against OpenAI in the USA and UK, starting at around 09:00 am ET. A look back at previous days shows this time is a common issue, suggesting it's simply putting strain on ChatGPT at busy times.

The OpenAI status page confirmed 'increased error rates', 'degraded performance', and 'login issues'. If not being hit with laggy performance, there are whole outages.

Altman took to X and addressed these teething problems as he told disgruntled users: "We are getting things under control, but you should expect new releases from openai to be delayed, stuff to break, and for service to sometimes be slow as we deal with capacity challenges."


ChatGPT has just closed a $40 billion round of funding, which is the largest private tech deal on record. Considering the bank accounts should be booming, the current issues were too much for many, who vented their frustrations and asked what they're paying for.

One angry user said: "Why did you roll out image generation to free users if you don't have the capacity and you are compromising the service for paying customers?"

Another added: "Have you considered removing the free tier? At this point, it’s despicable that pro users are subsidizing free users while nothing works for us."

Someone else concluded: "Excuses. How can you have this when you just received $40B - where is the money going to if not servers? Just makes more sense now why people need to install a local version of a competitor LLM like @deepseek_ai."

ChatGPT is back up and running at the time of writing, but with fears that this will keep happening as you flock to generate your own AI artworks, Altman continues to come under fire.

Choose your content:

an hour ago
2 hours ago
21 hours ago
  • Bloomberg / Contributor via Getty
    an hour ago

    Inside 'ridiculous' AI startup that Elon Musk placed $60,000,000,000 'gamble' on

    The company can make use of SpaceX's supercomputer that's powered by 200,000 Nvidia GPUs.

    News
  • d3sign via Getty
    2 hours ago

    Man with Instagram name ‘ihackedthegovernment’ receives startling sentence after pleading guilty to hacking the government

    Maybe things aren't as clear as they seem

    News
  • Bloomberg / Contributor via Getty
    2 hours ago

    How to get your money from Capital One's $425M lawsuit impacting 5M customers

    Capital One's switch from its 360 Savings account to its 360 Performance Savings in 2019, was listed as being deceptive

    News
  • Spencer Platt / Staff / Getty
    21 hours ago

    Grindr set to launch $500 AI feature as they host first White House Dinner party

    The popular hookup app has vowed to become an AI-first company.

    News
  • Warning to anyone using ChatGPT for medical advice as new study reveals disturbing results
  • OpenAI releases new ChatGPT that's more ruthless and has fewer restrictions
  • ChatGPT down for thousands across the world in huge outage
  • Sam Altman reacts to viral video showing ChatGPT hallucinating