HomeAI News
OpenAI CEO says custom GPTs delayed due to heavier-than-expected usage

OpenAI CEO says custom GPTs delayed due to heavier-than-expected usage

Hayo News
Hayo News
November 10th, 2023

Update: Weds. Nov. 8, 11:21 pm ET: In a post on X, OpenAI developer advocate Logan Kilpatrick said the company was experiencing “an abnormal traffic pattern reflective of a DDoS attack. We are continuing work to mitigate this.”

It’s been just two days since OpenAI unveiled its newest services to the world at DevDay, its first-ever developer conference in San Francisco, among them — a GPT Builder that enables third-parties to easily create their own customized, simple chatbot models for completing specific tasks atop OpenAI’s ChatGPT, as well as AI Assistants that can plug into outside apps use intelligence in them from OpenAI’s GPT-4 model, and new, reduced pricing for many of its tools. Oh, and a new, faster, GPT-4 Turbo model.

While the enterprise software community at large is still digesting the flurry of announcements, it turns out that the demand for these features has already exceeded what OpenAI anticipated: OpenAI CEO Sam Altman said today in a post on the social network X that the staggered rollout of GPTs in particular, originally scheduled to be available for all GPT Plus and Enterprise subscribers on Monday, November 13, has been delayed due to higher-than-anticipated usage of the company’s new tools.

In addition, Altman warned of “service instability” due to the demand load of many new users pinging the company’s servers for access to its models and new tools.

That tracks with what some users reported on X today, noting that service to ChatGPT in particular was interrupted.

While Altman did not provide any updated details on the timeline for when the GPT Builder and custom GPTs would be made broadly available, attendees at DevDay and other selected users were granted early access.

Some of them have already started building interesting new custom GPTs to perform such tasks as making original GIFs and product prototype imagery using OpenAI’s DALL-E 3 image generator model baked into ChatGPT.

Even though the increased usage of these new features is resulting in a delay of OpenAI’s release plans, it ultimately seems like a good problem to have for the company, showing that people are clamoring for what it has to offer.

Reprinted from Carl FranzenView Original


no dataCoffee time! Feel free to comment