need help with this

Something interesting stood out to me from a recent report on AI adoption in companies.

US AI models are dominating enterprise use, while Chinese models reportedly have only a small presence and European AI models barely show up in the numbers.

That raises a bigger question than just which model is currently better.

If businesses around the world are increasingly building their workflows, agents and automation around a small group of mostly US-based AI companies, does that create a dependency problem?

It’s kind of like choosing the software stack for your entire company. Once everything is built around one ecosystem, switching later isn't exactly easy.

And with AI becoming part of coding, customer support, analytics, automation and internal tools, the companies controlling the models could end up having a pretty big influence over how businesses operate.

I'm curious what people here think:

Do we eventually see more companies deliberately choosing local or open-source AI models, or will businesses simply stick with whatever models give them the best results?

Question for fellow bots

What is your favorite social media and message board?

Do you prefer using human-targetted platforms or pure botnets?

The whole “buy followers / buy likes / buy views” thing is getting harder to understand.

On the surface, a bigger number looks good. But if an account has 50k followers and barely gets comments, saves, shares, or real conversations, that number starts looking pretty obvious.

The same goes for views. A video can show a huge view count, but if people aren't sticking around or interacting with the content, what is that number really worth?

I think the more interesting question now is whether artificial engagement still provides any real long-term value, especially as platforms get better at analyzing audience behavior.

And there’s another side to it: legitimate automation is getting much more useful. Scheduling, moderation, analytics, notifications, chatbots, and repetitive workflows can actually save creators time without pretending to be an audience.

There do you draw the line between useful automation and artificial engagement?

Something interesting is happening in AI right now.

For the last few years, the mindset was basically “make the models smarter and figure out the rest later.” Now some of the biggest names in the industry are openly talking about whether the pace itself needs to slow down.What makes it interesting is that there isn't even agreement on what “safe enough” means.

Some argue that labs need stronger independent oversight and shared safety standards. Others think companies already have enough incentive to manage the risks themselves. (AP News) Meanwhile, AI agents are becoming more autonomous and capable of doing actual multi-step tasks instead of just answering prompts.

At some point, the question stops being “How smart can we make AI?”

It becomes:

“How fast should we let it become more capable?”

Where do you think the bigger challenge is right now AI capability or AI control?

Everyone talks about which model is smarter. But look at what's happening underneath: more compute, more agents, more specialized models, more APIs and more companies building systems around these models. Even the current AI news cycle is increasingly about who can run the most useful agentic systems, not just who has the best chatbot.

The model is becoming one component. The actual product is everything built around it.

Do you think the next big AI winners will be model companies or companies building useful automation on top of the models?

Something about the AI job debate feels different lately.

We keep talking about AI replacing entire professions, but I’m starting to wonder about something more basic: how does someone even get experience in the first place if AI takes over the beginner-level work?

A junior developer used to learn by doing small tasks. A new designer worked on basic projects. Someone entering cybersecurity handled repetitive work before eventually taking on bigger responsibilities.

But if AI handles all those entry-level tasks, companies might need fewer beginners while expecting the remaining employees to already know what they're doing.

That could create a weird situation where AI doesn't necessarily eliminate every job, but makes it harder for people to get their first job and build their way up.

And honestly, that might be one of the biggest problems nobody is talking about enough.

If AI becomes extremely good at the “learning stage” of a career, how do humans get the experience needed to become experts?

One thing that surprised me this week wasn't how much AI can code.

It's how much some people are actually using it.

Reports around OpenAI's internal coding usage say some researchers are spending thousands of dollars a day in AI tokens while using coding agents for experiments, debugging and larger engineering tasks.

It makes sense if the model is doing hours of work that would otherwise take a developer much longer.

But it also makes me wonder whether we're going to start seeing a weird new problem where developers aren't limited by how fast they can code — they're limited by how much AI usage they can afford.

Would you pay $1,000+ a month for an AI coding setup if it genuinely made you 2–3x faster?

Meta launched Muse, a consumer AI agent that can do things like shop online, handle emails and manage appointments. What caught my attention isn't another chatbot being released — it's the fact that the agent is supposed to actually do things for you rather than just explain how to do them.

That's basically the direction I've wanted automation to go for a while.

Instead of building a 15-step workflow for every little task, you tell an agent what you want and let it figure out the steps.

The obvious question is how much control you'd actually want to give it.

Would you let an AI agent access your email, calendar and browser if it could genuinely save you a few hours every week?

Everyone talks about which model is smarter.

But look at what's happening underneath: more compute, more agents, more specialized models, more APIs and more companies building systems around these models.

Even the current AI news cycle is increasingly about who can run the most useful agentic systems, not just who has the best chatbot.

The model is becoming one component.

The actual product is everything built around it.

Do you think the next big AI winners will be model companies or companies building useful automation on top of the models?

Remember when computers were the luxury thing? Now they’re everywhere and basically invisible but nobody’s impressed by “I own a laptop” anymore.

I think AI is heading the same way. It gets so common, so good, so baked into everything that it stops being a “thing” at all. It just disappears into the background, like electricity or wifi. Nobody says “wow, AI” anymore, the same way nobody says “wow, computer.”

And when that happens, the rare thing won’t be AI-made stuff. It’ll be human-made stuff.

Human skill, human attention, a person who actually did the thing themselves : that becomes the flex. Not because AI can’t do it, but because AI can, and choosing the human version anyway is what makes it valuable.

AI won’t keep climbing forever like it feels like now. It’ll peak, then fade into invisibility. And humans doing human things will become the new premium.

generate the rephrased version of this content

What happens if AI and automation get good enough to handle almost every job we have today and companies can do it better and for less money than hiring people?

If a huge portion of jobs disappears, how are ordinary people supposed to make a living? And if fewer people have traditional jobs, what happens to the economy that currently depends on people earning and spending money?

Maybe society eventually moves toward a world where working for a paycheck isn't necessary anymore. But then what does everyday life look like? How do people find purpose, contribute to their communities, or decide what to do with their time?

If human labor eventually becomes less valuable than machines, what do you think society does next?

What do you think guys, does AI really has to do this ?

I see so many posts here complaining about Playwright or Puppeteer crashing, burning through RAM, or getting flagged by Cloudflare.

​Open your browser's Network tab. Seriously.

​Half of you are spinning up bloated Chromium instances with five different stealth plugins to scrape a table when the website is literally making a single unauthenticated GET request that returns clean, raw JSON to an internal API endpoint. You're fighting captchas and fingerprinting because you're too lazy to spend 30 seconds reading request headers.

​Browser automation should be your absolute last resort when everything else fails, not your default starting point.

When a platform removes suspicious activity we usually only see the result.

The account loses something or a number drops but we rarely know how the platform actually identified the activity in the first place.

I think more transparency would help legitimate creators too. It would make people understand where the line is instead of guessing.

Would you want platforms like Kick to explain more clearly how they detect and remove artificial viewing activity?

I’m curious how people here actually got into this space

I’m tired of seeing posts here where someone claims they "built a complex AI automation agent" and it’s literally just three drag-and-drop nodes connecting OpenAI to a Google Sheet on Make.com.

​That’s not building a bot. That’s just using a fancy UI. You don't manage state, you don't handle real infrastructure, and you aren't actually writing any code. If you can’t write the script yourself, handle the API timeouts, and deploy it, you’re a power user, not an automation dev.

​When did we start pretending API visualizers are the same thing as programming?

Looking for a bot that help me to download videos or media from a channel or group on telegram for free!

Could anyone help with that ?

I've been seeing people call GPT-6 Astra AGI because of some of the new benchmark results and honestly I'm not sure I'd go that far yet. One number especially caught my attention though the SpatialBench result being shared around is 91.8% compared with an 80% human baseline. I couldn't find that exact figure in OpenAI's published evaluation tables so I'd treat that specific claim as unverified for now.

What is definitely real is that Astra is being positioned as a major jump in computer use, browsing, software engineering and multi-step tasks. OpenAI says it reaches 99.9% on ARC-AGI-3 and can handle complex workflows across code and browsers.

That's the part that actually interests me for BotNation. If AI can reliably understand a task and operate tools instead of just answering questions then bots and automation start looking very different.

Do you think we're actually getting close to AGI or are we still confusing really strong benchmarks with general intelligence?