Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Unexpected responses from ChatGPT: Incident Report

status.openai.com

Post reply on HN
Print view View original

271–277 of 277 posts

Re: Unexpected responses from ChatGPT: Incident Report

#271

Post by polynomial »

atleastoptimal wrote: post #17

Almost a tautological PR paragraph: "The words were wrong because the model chose the wrong words"

The fundamental, absolutely fundamental problem here is the assignation of agency to the model. This is milestone. Post-moderms are conducted when corporations need to assess responsibility and assign blame. Truly they have earned their ML generated MTG card status as mage to have right under the noses of everyone, shifted the very possibility of blame onto the "AI" giving it agency as a legal expediency, fiduciary if we're stretching it.

Re: Unexpected responses from ChatGPT: Incident Report

#272

Post by renonce »

jstarfish wrote: post #267
renonce wrote: post #191

Earlier quoted context omitted.

That sounds like a good use case for GPT. A GPT that automatically highlights such corporate speak and hints “WARNING: bullshit ahead”. I’m 100% sure it’s technically very easy to engineer such a model. Do you think OpenAI’s superalignment will ever allow you to make such a model?

You don't need a whole fucking model for this, you just need to invoke the right arcana. Try: > Act as a professional skeptic. Assess the passage for any deceptive, manipulative, or dishonest elements, including instances of doublespeak, inconsistency, fraud, disingenuity, deception, or sophistry: It works on ChatGPT 3.5 as long as you're not trying to ask it questions about the Talmud. That will quickly devolve into…

That prompt looks great. I tried it on OpenAI's incident response and got: (only last paragraph taken for brievity)

> ... > In conclusion, while the passage does provide a basic account of the incident, its lack of detail, reliance on jargon, and vague explanations leave room for skepticism regarding the severity of the issue, the transparency of the response, and the effectiveness of preventative measures. A more thorough and less technical explanation could have mitigated some of these concerns.

Cloudflare's postmortem: (https://blog.cloudflare.com/cloudflare-outage-on-june-21-202...)

> In summary, the passage does not exhibit overt deception, manipulation, or dishonesty. Cloudflare's statement seems to be a forthright attempt to explain what happened, accept responsibility, and outline steps to prevent future incidents. The inclusion of technical details and the acknowledgment of their error reflect a commitment to transparency, even if the promotional content at the end might seem slightly incongruent with the preceding apology and explanation.

Now the thing is to make a fast and reliable plugin that automatically puts the warning on the webpage and shame these corporates

Re: Unexpected responses from ChatGPT: Incident Report

#273

Post by joeythedolphin »

nottorp wrote: post #234
joeythedolphin wrote: post #225

Earlier quoted context omitted.

I don't know, it's more then this. I ask ChatGPT to teach me all about Hinduism, the siva purana, ancient indian customs, etc., and it is an incredible tutor. I can ask it to analyze things from a Jungian perspective. I can ask it to consider ideas from a non-typical perspective, etc. It is pretty incredible, it is more then a word gargler.

"teach you all". So you don't know enough to evaluate the answers. Or where to find the original sources it quoted or mixed.

Um yes I do? Lol why do you just assume this?

Re: Unexpected responses from ChatGPT: Incident Report

#274

Post by joeythedolphin »

EchoChamberMan wrote: post #236
joeythedolphin wrote: post #225

Earlier quoted context omitted.

I don't know, it's more then this. I ask ChatGPT to teach me all about Hinduism, the siva purana, ancient indian customs, etc., and it is an incredible tutor. I can ask it to analyze things from a Jungian perspective. I can ask it to consider ideas from a non-typical perspective, etc. It is pretty incredible, it is more then a word gargler.

It is, unquestionably, from the math, a "word gargler" as you put it. https://xkcd.com/1838/

No it's a digital neural network expressing it's inputs and outputs in language.

Re: Unexpected responses from ChatGPT: Incident Report

#275

Post by okdood64 »

troupo wrote: post #192
okdood64 wrote: post #10

Say what you will about Google, but bugs such as these, released to the public (or their enterprise customers) this debilitating to a core product, are exceedingly rare.

I see you haven't seen the Gemini debacle: https://twitter.com/alexthechick/status/1760331428133253312

This isn't really debilitating. Yes, it's a big "bug", but that doesn't make the product unusable.

OpenAI effectively had a global outage on ChatGPT.

Re: Unexpected responses from ChatGPT: Incident Report

#277

Post by EchoChamberMan »

joeythedolphin wrote: post #274
EchoChamberMan wrote: post #236

Earlier quoted context omitted.

It is, unquestionably, from the math, a "word gargler" as you put it. https://xkcd.com/1838/

No it's a digital neural network expressing it's inputs and outputs in language.

If you trust the words that machine strings together, you're making dangerous assumptions - look at the lawyers getting busted for submitting fake cases. These things are literally designed to sound plausible as the primary concern - facts aren't really a thing.
Post reply on HN