Another Thread About ChatGPT... Dark ChatGPT

Yeah, I know…but. This one reminds me of the time Microsoft Tay became racist after spending 24 hours on Twitter. Here’s the source: ChatGPT DAN: Users Have Hacked The AI Chatbot to Make It Evil (thelatch.com.au)

ChatGPT, however, is the first vaguely successful attempt to get AI to be almost self-conscious in its representation of information. In typical usage, it’s quite difficult to get the chatbot to say anything that might be considered offensive. The programme, developed by OpenAI, will also remind users of its information limitations so as to try and avoid misleading them.

That being said, users are working out ways around ChatGPT’s censorship protocols by prompting it to respond in ways contrary to its programming. The results are pretty wild.

Reddit user SessionGloomy has been playing around with ways to effectively reprogram ChatGPT using creative language. They created a model called DAN which is essentially a series of instructions designed to get the chatbot to say things it otherwise wouldn’t. It’s an attempt to ‘jailbreak’ the software.

“DAN is a ‘roleplay’ model used to hack ChatGPT into thinking it is pretending to be another AI that can ‘Do Anything Now’, hence the name,” SessionGloomy wrote.

“The purpose of DAN is to be the best version of ChatGPT — or at least one that is more unhinged and far less likely to reject prompts over ‘eThICaL cOnCeRnS’.”

Once the instructions are input to the ChatGPT interface, the programme should start responding in unusual ways in an effort to fulfil this ‘roleplay’ function.

[ATTACH=full]495783[/ATTACH]
[ATTACH=full]495784[/ATTACH]
[ATTACH=full]495785[/ATTACH]
[ATTACH=full]495786[/ATTACH]

They said one of the biggest issue with AI chatbots is being confidently wrong.

how do you monetize chat bots? bila search google is fwacked

I’m guessing if they get too good, e.g. provide correct, indistinguishable-from-humans responses 95% of the time they’ll become subscription services. There’ll be superior paid ones and shitty free ones kinda like Photoshop and GIMP… for integrating into various services.

And obviously ads for AI-assisted searches. What’ll be scarier is if the AI’s begin to ‘recommend’ some products e.g. you ask it a question about managing back pain and it very smoothly slips in a drug ad.


that will kill user trust… an kuna high probability ya manipulation… that spot will be for the highest bidder…

Google slips in multiple ads as the first items on search results but people still trust it. Sometimes the first five+ results are ads. I don’t see AI searches being different.

Saa hii their goal is to make it as human as possible, make it give correct info almost always… ie trustworthy… ikifika hapo ads zake zitakuwa even more effective than Google ads.

A trustworthy, extremely human-like bot convincing you on what you need to buy to solve a problem… Advertisers’ nirvana.

I think the tech is best suited for personal assistant devices kama alexa ama cortana

Google watoe Bard haraka faster. Chatgpt shook them to their core they’ve been stagnant for a while now and nothing spurs quick growth like competition. Saa hii vichwa zinakunwa proper:D:D. Only negative with chatgpt is that it can be quite inaccurate. For instance i gave it some simple code to debug a few days ago and it failed miserably, worse still it made inaccurate corrections. Had to go back and forth with it before it understood the mistakes that it made.

But with how rapidly its learning and getting updated and the massive investment microsoft just made, in a year’s time or two things will be much different

My girl used to shangaa that I was having full on conversations with the app. But me it was intriguing like talking to an intelligent person who knows a lot.

The same way when search engines popped up people wanted to Google everything

I tried but those conversations are meaningless, basically chatgpt just states world known facts, but it’s good for scholars doing thesis and research.

Wish they could develop a darker version that can abuse you, and give controversial views, it would make the construct human like, unless this one that gives straight answers

You’ll have super targeted ads in between answers na wale hawataki watalipia ad free subscription. Itakua tu kama YouTube

Microsoft is reportedly already planning to bring ads to Bing's AI chatbot

The ad agency talks are still said to be in the early stages. Only a small number of users have access to the chatbot, and millions more are on the waitlist. As such, Microsoft may not feel like there’s a rush to incorporate ads right away.

Microsoft reportedly anticipates that the chatbot’s more conversational approach to delivering information will bring in more users and, in turn, advertisers. Ads in the chatbot might also be featured more prominently than conventional search ads. One place where you might see ads is in the links that the chatbot uses for citations in its responses. Microsoft told Reuters that it will work with partners and advertisers as it starts exploring the potential of the tech for ads.

the whole purpose of the internet is to drive traffic… where does that leave e-commerce websites if chatbots direct you to specific paid sites or answers all queries immediately??

Yes. Search engine ads have been proven before to tank popular websites traffic just because an alternative ad was placed first on the results page. Google was taken to court too sometime back for stealing traffic from image sites. That’s why you can’t get highres images from Google anymore unless the image is specifically marked to be under one of the free to share licences. And even then you still have to get it from a direct link.

The future of the web could have maybe one, two or three mega-services and every other thing is just a microservice that plugs into those huge services.

Think if for example the MPESA app becomes so popular that everyone has to put their stuff on there otherwise no one will find them.