News for hackers. Stuff that matters.
Re:Latest ChatGPT 4 System prompt (1,700 tokens) by iambateman on Wednesday February 07, 2024 @03:45PM (#39294080)
I don’t know…if I watch a commercial, I know there were guidelines given by the company that back their values.

Chick-fil-A employees are trained to say “my pleasure” but customers don’t watch their training video.

I can appreciate that ChatGPT is a product, built by people with values, and this is their way of inserting those values.

Re:Latest ChatGPT 4 System prompt (1,700 tokens) by eli on Wednesday February 07, 2024 @04:22PM (#39294484)
I think the goal is to have a less obviously racist product. This is a business and PR concern, not a moral one.
Re:Latest ChatGPT 4 System prompt (1,700 tokens) by slingnow on Wednesday February 07, 2024 @04:01PM (#39294281)
Please elaborate on how watching a commercial or receiving a pleasantry from a cashier is anything remotely like a company secretly influencing the results of a search for knowledge.
Re:Latest ChatGPT 4 System prompt (1,700 tokens) by Fripplebubby on Wednesday February 07, 2024 @04:08PM (#39294342)
Your "search for knowledge" occurs through the portal of OpenAI's ChatGPT software, which is a consumer-facing product just like a commercial and just like customer interactions at a shop, and so if we (society / the law) do not question and regulate commercials and customer interactions in this way, then we also should not question or regulate OpenAI's system prompts, since these mechanisms are so similar.

If you want an un-influenced "search for knowledge", you are well within your rights to pursue that independently using your own model and software, but because you are accessing software developed by another company, these rights do not apply.

Re:Latest ChatGPT 4 System prompt (1,700 tokens) by singleshot_ on Wednesday February 07, 2024 @04:48PM (#39294735)
This is exactly the same justification for why social media platforms must retain the ability to shape the content in their platform through editorial decision making. Both types of firms sell a carefully created coherent speech product and must be allowed to retain the right to make commercial decisions to shape their customer base.

Don’t like it? find another platform. Pretending this is censorship and not commerce is wrongheaded.

Re:Latest ChatGPT 4 System prompt (1,700 tokens) by Sabinus on Wednesday February 07, 2024 @06:39PM (#39295835)
Would you consider the highly curated experience of the HN front-page to be censorship? Mod actions and anti-drama algos effect your experience here.
Re:Latest ChatGPT 4 System prompt (1,700 tokens) by supriyo-biswas on Thursday February 08, 2024 @03:54AM (#39299806)
I guess social media is generally not claimed to be a curated experience, whereas on HN the curation is explicitly described with words like “gratifying intellectual curiosity” and the like.
Re:Latest ChatGPT 4 System prompt (1,700 tokens) by singleshot_ on Thursday February 08, 2024 @04:46PM (#39308170)
I would consider it to be a coherent speech product that HN/YC has just about every right to curate as they see fit to match their commercial goals.

So many people disagree on what censorship is that I prefer to be more precise when possible.

Re:Latest ChatGPT 4 System prompt (1,700 tokens) by nostromo on Wednesday February 07, 2024 @04:02PM (#39294285)
Your iPhone is product - what if Apple decided you shouldn't be able to view adult material on it? Or what if Gmail decided you shouldn't be allowed to use certain harsh language in emails you send?

Where do you dry the line personally? I find the idea of corporations and SV tech bros trying to define my values repulsive.

In a competitive environment I could simply choose another service provider - but most of these tech giants are monopolists of one variety or another.

Re:Latest ChatGPT 4 System prompt (1,700 tokens) by 19h on Wednesday February 07, 2024 @04:26PM (#39294528)
Pretty sure that’s exactly what Steve Jobs decided [0] :-)

[0] https://www.wired.com/2010/04/steve-jobs-porn/

Re:Latest ChatGPT 4 System prompt (1,700 tokens) by smith7018 on Wednesday February 07, 2024 @04:31PM (#39294573)
Then don't use ChatGPT. There are hundreds of other models and ways for you to use an LLM without OpenAI's injected prompt.

> I find the idea of corporations and SV tech bros trying to define my values repulsive.

They're not. They're reflecting _their_ values in their product. You're not entitled to use their product with _your_ values.

Re:Latest ChatGPT 4 System prompt (1,700 tokens) by lencastre on Wednesday February 07, 2024 @05:04PM (#39294902)
Esteban Trabajos really said that the consumer doesn’t know what it wants, the consumer must be shown the jesus phone and realize it wanted that phone all along. And if copy paste wasn’t possible on the iPhone 2G or apps couldn’t run in background that’s because ol’Stephen Works didn’t want it just yet. One day somebody will unlock the VisionPro HubPr0n and then and only then will VP become the de facto standard AR/VR headset approved by the pr0n industry, as Steven Efforts intended.
Re:Latest ChatGPT 4 System prompt (1,700 tokens) by stubish on Thursday February 08, 2024 @08:59PM (#39310516)
I find the Declaration of Human Rights is always a great starting point. Society is slowly forcing corporations in that direction.