Who is using ChatGPT?

[PaulEchere]

Ah, well I personally don't do a lot of (semi-complex to complex) coding. The stuff I tinker with any of them can do (tables etc). So, a personal view won't be valid.

And, with the pace of development around different models, it'll change almost weekly for the foreseeable future and sometimes every few days (last week, for example, was manic).

But a recent study I saw indicated that Deepseek currently has the edge for coding (over GPT, Claude, and GROK latest models)

I've tried doing some side by side comparisons of ChatGPT and Deepseek and everything I have so far points to ChatGPT's superiority in terms of coding.

What I've been doing is I was giving GPT and Deepseek user stories of different detail and depth (like you would with a regular developer) and the output produced by GPT (specifically o3-mini and o3-mini-high) are way more wholesome compared to those produced by Deepseek. Al least with hte use cases that I had in mind.

Haven't tried GROK or Claude tho.

[Jasper616]

We regularly use ChatGPT, especially for our affiliate website. It has become an essential tool in our daily business operations. From content creation to brainstorming marketing strategies, it saves us a lot of time and effort. The possibilities are endless, and we keep discovering new ways to integrate it into our workflow!

[wonderpunter]

Great for ideas, but always need to double check it as the info is nearly always wrong

[CritiqueJeu]

It's great to see how everyone is using ChatGPT! I've been using it mainly for quick research and educational purposes. It's like having a research assistant that helps me grasp complex concepts or get historical insights in no time. Super handy for projects or presentations!
However, I totally agree with the need to double-check the facts it provides, especially for critical tasks. It's a powerful tool, but it's always good to confirm the details.

[PaulEchere]

An update to some of the earlier chatgpt coding experiments that I've posted about here earlier.

I now have an odds comparison page on one of my websites - https://www.topbettingsites.ng/compare-odds

everything to run this page, scripts, scrapers, databases, uis and various bits of related matching and sorting logic were all created by chat gpt.

[jasminholt]

GROK has the best sense of humor
Claude is great for coding
Perplexity is best for research
GPT to refine your content

[Gypsy 1win]

Appreciate all the insights here! Personally, I’ve found GPT (especially the o3 variants) to be more dependable than Deepseek for the kind of structured, goal-oriented outputs I need, especially when it comes to crafting workflows, refining messaging, or handling semi-complex logic. That said, I believe every AI model - whether it’s GPT, Deepseek, Claude, or even GROK - needs a solid background check before trusting any output blindly. We use AI regularly in our affiliate marketing workflow, but only after validating facts or cross-checking strategic recommendations. The speed and flexibility are unmatched, but accuracy still needs that human layer. For me, GPT strikes the right balance between reliability and versatility. Deepseek has promise, but in real use cases, GPT just performs better.

[BalticR]

Yeah, ChatGPT has been in regular use, however, maybe more in other stuff than affiliate marketing.

[casinobonusguy]

I tried it for the first time this week and if I were a writer I would be extremely worried as reviews I paid for in the past were not as good quality as CHATGPT .I even ran it in my Copyscape and was truly unique.

[xecutable]

Great for ideas, but always need to double check it as the info is nearly always wrong

This exactly. And if you wanna toy with the AI's intelligence just tell it "you are wrong" and watch it change its answer to 180 degrees. Ive done some programming recently in Assembly x86 and had a buffer overflow, simple security flaw, easily fixable. ChatGPT couldn't fix it in 18 tries, and we are talking about 20 lines of code.

Ive asked it about C++ stuff, it gave me a bunch of libraries with such confidence that I took them and used them, just to understand they were deprecated 2 YEARS ago. With access to the internet, why it couldn't see that?

My sister was showing me a bunch of "legal requirements" in the US, citing ChatGPT. I told her to just write "you are wrong" and watch what happens. Within seconds the AI went from YES THIS IS A LEGAL REQUIREMENT, to NO IT'S NOT LEGALLY REQUIRED.

For an industry that drives stock prices up and consumes billions and billions of investments, this is a big sham to me. Yes it can rewrite small 2 liners in a perfect way, or some one paragraph bio for your Linkedin, but for highly complex and precise things, or to find the right answer? You better know the right answer to get the right answer.

So when I hear people be like "This will replace ITs very soon" my response is always "You don't use AI much do you", cause those who use it daily for wide variety of tasks, know how wrong it is.

[Strider1973]

AI has gotten much better.

It's an incredible tool and a lot of jobs will be replaced.
Also its graphics design quality has improved enormously since the beginning of the year.
I think many PC related task that were outsourced or done by specialists can now be done by ChatGPT (or its competitors).

Just in our industry from my experience:
- create an image instead of buying a stock photo + having to write where you bought it
- generate article ideas + write the articles instead of dealing with a writer
- doing translations
- fix errors on WP website instead of googling for the error code and hoping someone had the exact same issue before and posted about it

I just read about people using Upwork as freelancers that the earn much less this year compared to the previous year. Guess why... it's exactly that kind of jobs that get replaced by AI first.

On top of that, I use it daily to have an "expert" to chat with about any subject of private interest (asset allocation discussion, city trip travel planning, finding a new curved desktop monitor). Part of it are things that I wouldn't have researched before, but much of it is stuff that I would have used Google for in the past, and the quality of the results/answers is much better and much faster.

It really amazes me that people still aren't using ChatGPT or another AI...

[heikoritter]

I often use GPT in my job. This AI sometimes give not true answers

[slotsdudes]

Who is using ChatGPT and what is your experience.

I have started to use quite a bit however mainly for research, how to etc, and really enjoying it. Of course ChatGPT can be used for all kinds of great things but for me ... slow steps. lol

Recently bought a new iPhone 16 Max Pro and had to migrate from Android. ChatGPT saved me hours of searching on the internet "how to".

So far so good ... next is to dig deeper into analysis of spreadsheets.

Does it work well with spreadsheet? Chatgpt paid version could barely read a dict in JavaScript of about 150 entries without messing up

[Kanom Media]

Our competitors use it, mainly spinning our hard work through it and passing it off as their own.

[Sherlock]

This exactly. And if you wanna toy with the AI's intelligence just tell it "you are wrong" and watch it change its answer to 180 degrees. Ive done some programming recently in Assembly x86 and had a buffer overflow, simple security flaw, easily fixable. ChatGPT couldn't fix it in 18 tries, and we are talking about 20 lines of code.

Ive asked it about C++ stuff, it gave me a bunch of libraries with such confidence that I took them and used them, just to understand they were deprecated 2 YEARS ago. With access to the internet, why it couldn't see that?

My sister was showing me a bunch of "legal requirements" in the US, citing ChatGPT. I told her to just write "you are wrong" and watch what happens. Within seconds the AI went from YES THIS IS A LEGAL REQUIREMENT, to NO IT'S NOT LEGALLY REQUIRED.

For an industry that drives stock prices up and consumes billions and billions of investments, this is a big sham to me. Yes it can rewrite small 2 liners in a perfect way, or some one paragraph bio for your Linkedin, but for highly complex and precise things, or to find the right answer? You better know the right answer to get the right answer.

So when I hear people be like "This will replace ITs very soon" my response is always "You don't use AI much do you", cause those who use it daily for wide variety of tasks, know how wrong it is.

I am very sceptic towards AI, but slightly in a different way. Most people are simply dumb and will not be able to squeeze the potential out of it. The same happened with crypto, social networks, IoT or internet generally. People are, me included, simply too lazy. Changes occur slowly. For example e-government, is being deployed now in most countries. There was HW and SW in 90s. Just there was not the will.

People will use the AI as smarter Google, as many do it, but AI is poor at googling. At least by now. It is not something intelligent, it is just an autistic language model. Anyone thinking that it is more is crazy. Being able to search for information, to avoid deception, is one of human abilities (at least by now).

AI for most people means and will mean directly a little. Yes, they will learn to summarize info faster; maybe with time they will see that random search results are worse than from google. But that does not support the hype that is around it, at least not in a way it is perceived.

But, I must admit I was too much affected by the hype and obvious idiocy around it, that I refused AI for myself and it was a mistake.

On a personal level, AI is for me a revolution, that I did not expect to happen. Just few days ago I bought / rented Claude after they released new code and only my first LSD trip and my first money made online can match the experience. Since the day I started with it, I am just sitting here and working. I was able to do things I was sure I will not be able to do in years in those days. The more I work, the more I actually need to work, because the finished tasks open path to new tasks that I need to catch before I die.

I am not using it in a classic way (using Putty), the setup is pretty unintuitive and inconvenient and it is a result of AI work of my collaborator, who adjusted it.

It writes code. Faster than anyone by some orders of magnitude. Indeed it makes incredible mistakes, both stupid and dangerous. I never programmed and my life was dependent on those fachidiots. No more. After I started to play with design of my future workshop, which frankly is still better to draw in AutoCAD (probably until I get more experience), I set up quite complicated network of many parallel routers that somewhat worked, but did not do what I wanted (I was not able to access diesel generator or other network points from other parts of the network. I not only set it up, I even learned how it works. Then I started to redo some lending bot, for now just the part that has read access and notifies me on telegram, which is crucial for revenue, because I can not be logged in all the time. After that I started to play with relay from Cerbo GX from my fotovoltaics, knowing absolute **** about it and in one day I set up Node Red flow that has logic which switches on and off my electric water heater based on logic how much excess of solar energy I have and how much free power has the invertor. After that I understood that this is not a limit and I started to add closed source Tuya smart switches to my electric network with plan to command them also with the FVE logic, which showed like a task that not many people achieved (maybe no one). That AI thing helped me to crack the HW and it works. We were stuck 10x, solving same error over and over. When the unstoppable thing was giving up, always a bit of human googling helped to bring new ideas or come out of "mental" circles. Now I am working on additional logic, UX; also on different projects at the same time. Thank god there the expiring tokens and I will not buy more expensive plan, otherwise I do not sleep, eat, go out etc. All you have to do is to have a goal and not give up the questions and demands. Then even if you know nothing about the topic, you can be on good enough level in everything quite fast. But expect lie or problem in every byte of info that is coming from it.

I am pretty sure most people will not try to do those things, so life will be the same. Just more grey, less work, less salary. Step by step. People will be dumber from chatbot generic responses and it will affect them paradoxically in a way that they will not be able to give the right prompts to AI.

Few years ago, I saw a poster, where old miners got money to requalify to programmers. Hopefully they will make it to pension. I think they will, the changes are slow. But there is an option for personal revolution at least. Buying Ai stocks companies or betting on it is tricky. Either it will be omitted and abused in wrong way like all nice inventions and it will not make sense. Or it will prevail with time and then companies, humanity, social adherence... that all will vanish, because it will be like with SERP. Winner will take it all; then in reaction to AI feudalism only more repression will come, but the totality will come from the dumb outsiders, because insiders will be in.

I would like to write something more about philosophical bs, but understand me, I have more exciting things todo...

[gregormlakarrr]

Yes, I use it, mostly for content. If you learn more and start using the right promos, your content will reach another level.
I have a question: maybe someone can recommend a good AI for audio?

[Max West]

I’m using ChatGPT mostly for research, quick explanations, and step-by-step “how to” help. It’s been especially useful when switching devices — like moving from Android to a new iPhone — saving a lot of time compared to searching everything manually. Next, I want to explore its use for spreadsheet analysis and data insights.

[GlobalPokerSites]

Use ChatGPT daily and whilst great it often gives outdated/wrong info.

[allfreechips]

I use chat for answering a question i have, its actually quite well for this, and making images. Still google if i want to find something though.

[Teodora B]

I probably use it daily, mostly for research. It actually saved me recently when my laptop's wifi was not starting, and I managed to fix it with one of the solutions ChatGPT offered. I also use it sometimes before making a purchase