We have been gorging ourselves at an all-you-can-eat buffet, but the menu is about to change. AI tools that have been discounted will soon start charging based on your usage, raising serious questions about our appetite to pay.

The answer could have big implications for the future of artificial intelligence (AI).

And the AI Tech bros aren't doing too well in the stock market at present.

https://thepoint.com.au/opinions/261026-the-cost-of-using-ai-is-about-to-skyrocket-but-do-we-have-the-appetite-to-pay

all 16 comments

sorted by: hot top controversial new old
[–] 1 point 4 hours ago*

If experienced staff are pushed out because essential work is done by AI , it will be hard to go back – a discount can expire a lot faster than it takes to rebuild a workforce.

Looking forward to the downfall of companies that fired most of their staff, or didn't hire any, assuming an LLM would do the job.

  • source
  • [–] 1 point 11 hours ago (4 children)

    We will never pay for AI but there are millions of people who will. Downvote me all you like but it truly is inevitable

  • source
  • hideshow 4 child comments
  • [–] 5 points 10 hours ago*

    AI costs are currently heavily subsidised. Do you think millions of individuals will be willing to pay 100 AUD a month? 500 AUD per month? Some companies might, but I doubt individuals will.

    Downvote me all you like but it truly is inevitable

    Let's leave this shit behind on Reddit.

  • source
  • parent
  • [–] [S] 3 points 11 hours ago (1 child)

    Why should I downvote you (in any case, I hardly ever downvote anyone)? You may be absolutely right. I put up the post because I was unaware that the cost of using AI was going to go up even more and it was worth discussing.

    The idea here, I hope, is to discuss ideas and opinions. Sometimes, people make comments with malicious intent rather than comment on the topic. I don't even downvote them but I might give them a serve back.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 5 points 17 hours ago

    Think harder, AI

  • source
  • [–] 1 point 15 hours ago (5 children)

    for devs (ie the only part of the bubble that is seeing any significant usage), qwen 3.8 and deepseek are fine for %95 of use cases.

    there are a bazillion providers that will sell you tokens for peanuts. mix that in with self hosting where you have the resources to do so (its docker containers and a bit or orchestration not rocket science) and we will just see the market normalise to something closer to web hosting.

    ...and windows monkey IT departments will just buy whatever microsoft is selling.

    that future is not worth 1/1000th of whatever todays AI hype cycle is getting shoved down our throats.

    </pulling some shit out of my arse>

    services:
      llama:                                                                                                                                                             
        image: ghcr.io/ggml-org/llama.cpp:server-cuda                                                                                                                    
        container_name: llama                                                                                                                                            
        env_file:                                                                                                                                                        
          - .env.llama                                                                                                                                                   
        ports:                                                                                                                                                           
          - "8081:8081"                                                                                                                                                  
        expose:                                                                                                                                                          
          - '8081'                                                                                                                                                       
        volumes:                                                                                                                                                         
          - ./models:/models                                                                                                                                             
        networks:                                                                                                                                                        
          - proxy                                                                                                                                                        
        deploy:                                                                                                                                                          
          resources:                                                                                                                                                     
            reservations:                                                                                                                                                
              devices:                                                                                                                                                   
                - driver: nvidia                                                                                                                                         
                  count: all                                                                                                                                             
                  capabilities: [gpu]                                                                                                                                    
    
  • source
  • hideshow 5 child comments
  • [–] [S] 1 point 11 hours ago (3 children)

    Must be painful if that's what you pull out of your arse 🤣

    Can you explain in non-techwrite what you mean by your 'services' segment? Or am I intentionally not supposed to understand?

  • source
  • parent
  • hideshow 3 child comments
  • [–] 2 points 9 hours ago (2 children)
  • [–] 7 points 1 day ago
    [–] 5 points 1 day ago
    [–] 3 points 1 day ago

    They're about to learn a hard lesson

  • source