Has 'Chat GPT' proven helpful?

Post Reply New Topic

Serious responses only please.

Enclosed is a summarized Chat-GPT statement in WP Discussion - 'Chat-GPT Advice On Friendships Via Shared Activities.'

viewtopic.php?f=7&t=415552&p=9341158#p9341158

I used ChatGPT for responses to interview / job hunting email.

It made sense to me rather than google sample response letters and trying to edit them together and customize them for me situation- which is one of the things ChatGPT is good at.

How has 'Chat GPT' proven helpful in living with the Autism Spectrum?

Just as posted before, responses to job related activities is way easier with it. Jobs = money = way to sustain life.

It can help rewrite what I want to say to make it more socially acceptable. Example saying I left my last job because it was contract and will not be returning due me turning them into DOL is apparently a turn off. Focusing on positives like "I'd like to branch out" seem to be better. Why? No clue I just live here on the...wrong planet

Related WP discussion-thread (LINK) 'Life Hacks helpful with AS?'

LINK: viewtopic.php?t=415955

I don't rely on ChatGPT for anything, I do long based Text Based stories but I refuse to rely on AI for stuff

AI has been taking away my editing work so not helpful at all to me.

I've read a fair bit about ChatGPT but have never used it.

I'll report back here if I ever do end up using it.

I had asked 'Chat GPT' the following inquiry: What are the reasons why people are cynical about 'Chat Gpt?'

ChatGPT response:

People may be cynical about ChatGPT or similar AI chatbots for various reasons, including:

Lack of Human Understanding: ChatGPT can produce responses that seem human-like but lack true comprehension of context and emotions. This can lead to users feeling that the AI doesn't truly understand them.

Inaccuracies: ChatGPT is not always accurate in its responses, which can frustrate users seeking reliable information or assistance.

Bias: AI models like ChatGPT can inadvertently generate biased or offensive content, reflecting the biases present in the data they were trained on.

Ethical Concerns: Users may have ethical concerns about AI chatbots, particularly in contexts like therapy or counseling, where human empathy and ethical judgment are highly valued.

Privacy and Security: Concerns about data privacy and security can lead to skepticism, as users worry about the information they share with AI chatbots being mishandled or misused.

Job Displacement: Some people fear that the widespread adoption of AI chatbots could lead to job losses, particularly in customer service and support roles.

Unpredictable Behavior: ChatGPT's responses can sometimes be unpredictable or nonsensical, making it difficult for users to rely on the AI for consistent assistance.

Misuse: People are concerned about AI chatbots being used for malicious purposes, such as generating spam, fake news, or engaging in harmful trolling.

Lack of Accountability: Users may not fully understand how AI models make decisions, which can lead to frustration and mistrust when they receive responses they don't agree with or can't explain.

Technical Limitations: ChatGPT may struggle with complex or nuanced conversations, leading users to doubt its capabilities.

Fear of Dependency: Some individuals worry that an overreliance on AI chatbots could lead to a decline in human communication and problem-solving skills.

Hype and Misrepresentation: Overhyping AI capabilities in marketing and media can lead to unrealistic expectations, resulting in disappointment when users interact with the technology.

Disconnect from Human Touch: In situations where personal connection and empathy are essential, like healthcare or therapy, users may be skeptical of AI chatbots' ability to provide the needed emotional support.

It's important to note that while these concerns are valid, AI chatbots like ChatGPT have the potential for many valuable applications, such as providing information, automating routine tasks, and assisting with language translation. Addressing the above issues through continuous research, development, and responsible deployment is essential to improve the public perception and trust in AI chatbot technology.

The more Google gets gimped, the more I pose my questions to Chat GPT.

What is happening to Google because as far as I know, it's the same as usual. Lots of people including you keep saying it isn't though.

Yes.

I have a very specific example.

The company I work at has so far typically only had me do internal web apps where they'd only be available on the local network, hence jwt token protection on the API's and good passwords with 100K hashing was enough. They finally pushed me to build my first outward (open internet) facing web app and that's where I got the idea that the smartest thing I could do was build an OAuth 2.1 code flow with a dedicated login site for authentication / authorization, sending the antiforgery token to the app server via back channel, implementing PKCE, etc. for an Angular front-end and .Net Core back-end program (w/ .Net Core OAuth server). I can already feel the question coming up and the answer is no - I didn't get openid working (library issues) so I did this instead. I could maybe go back for that one of these days if it provides significant improvements but for now I had two projects - due within a month of each other - that were outward facing / public apps so I needed to get some completed security wrapper going that was up to the challenge of universal access.

It took about three weeks of beating on ChatGPT (as politely as I could - just a lot of questions), beating on code, more failing libraries, asking ChatGPT how I could replace those libraries with my own code, and finally after a pretty rough fight I finally got it all working.

Without ChatGPT being able to offer as much help as far as workarounds I probably would not have succeeded because there were perhaps a half-dozen items that would have been dealbreakers. As it stood the tutorials available online were either incomplete or didn't fit my use case, I got some ideas from some people like Raw Coding (Youtube channel) for what was important, past that when I said I was beating on ChatGPT I just mean those days where the code wasn't working, I'd give it to ChatGPT for review, ChatGPT couldn't find problems, so it was a matter of just stubbing my toes until everything fell into place.

Extremely.

I now use it as my primary source of information, and use Google as a fallback. I'm a paid subscriber, and only use GPT-4 (I've found GPT-3 - the version available to the public - to be on average less helpful than Google). As a software engineer, I have it write most of my code, and it's an incredibly useful learning tool. It's made learning fun for me again.

Has it been useful being Autistic?
Yes - I've found it very helpful when trying to figure out what other people mean, or trying to figure out what someone meant or might have implied when they said something. I will copy-paste snippets of conversations or messages that I don't understand, and it will explain the context, innuendo, and implications that I missed.

How do you get it to do that? I thought AI was even worse than autistic people at working out context, innuendo and implications.

editing

Who are you talking to? Me? What does your comment mean?