Fubarberry
@Fubarberry@sopuli.xyz
preview
This is a look into how an AI bubble burst might play out. The biggest issue is money, companies like OpenAI and Anthropic are burning money with no good way to turn it profitable. Google can “win” the AI race just by continuing to spend money, and wait for the others to crash and burn.
OpenAIs recent attempts to turn a profit (shopping and sora) are both failing, and they’ve resulted to trying ads (which they had previously described as a last resort).
Anthropic is capturing the developer market, but some reports say their metered models cost 5x more than what people pay for them. It doesn’t matter that they’re having market success, if more users just means burning way more money.
If multiple major players crash or have to get bought out, we may see several major datacenters shutdown or underperform, which could have a strong ripple effect on the economy.
Noteworthy, this article doesn’t include anything about the US government’s involvement in AI. The US considers AI a race to reach AGI before china, and may attempt to prop up a failing AI market to achieve this.
This is a look into how an AI bubble burst might play out. The biggest issue is money, companies like OpenAI and Anthropic are burning money with no good way to turn it profitable. Google can “win” the AI race just by continuing to spend money, and wait for the others to crash and burn.
OpenAIs recent attempts to turn a profit (shopping and sora) are both failing, and they’ve resulted to trying ads (which they had previously described as a last resort).
Anthropic is capturing the developer market, but some reports say their metered models cost 5x more than what people pay for them. It doesn’t matter that they’re having market success, if more users just means burning way more money.
If multiple major players crash or have to get bought out, we may see several major datacenters shutdown or underperform, which could have a strong ripple effect on the economy.
Noteworthy, this article doesn’t include anything about the US government’s involvement in AI. The US considers AI a race to reach AGI before china, and may attempt to prop up a failing AI market to achieve this.
preview
OpenAI also estimates that a similar percentage of users show heightened levels of emotional attachment to ChatGPT, and that hundreds of thousands of people show signs of psychosis or mania in their weekly conversations with the chatbot.
OpenAI also estimates that a similar percentage of users show heightened levels of emotional attachment to ChatGPT, and that hundreds of thousands of people show signs of psychosis or mania in their weekly conversations with the chatbot.
preview
Accurate weather predictions are one of the areas that I think AI has great potential. Traditional methods for predicting weather/storms is still very hit or miss, there’s a lot of interacting factors that make it difficult to accurately simulate. Machine learning is well suited to tasks that have lots of available data that’s complex to simulate accurately, and there’s real potential here I think.
Accurate weather predictions are one of the areas that I think AI has great potential. Traditional methods for predicting weather/storms is still very hit or miss, there’s a lot of interacting factors that make it difficult to accurately simulate. Machine learning is well suited to tasks that have lots of available data that’s complex to simulate accurately, and there’s real potential here I think.
preview
Hi, I’m Fubarberry, and I recently took over moderation of this community due to the original mod being inactive.
This community isn’t very active right now, but I wanted to talk to what members are here about what you want from this community.
For example:
- should this community focus more on AI in general, AI as a tool for creating things (images/video/music/stories/etc), or something else?
- Should this sub be focused on a more positive discussion of AI? Personally I feel like there’s no shortage of places on lemmy to doompost about AI right now, so this community would be more useful discussing beneficial uses of AI. Constructive criticism/acknowledging of issues is fine.
- Should we focus on AI tools available to end-users, and/or discuss AI news in general (AI used for medical research for example).
- Any other thoughts and opinions you have on this community.
Also side note, but right now this community has a bit of an issue with new posts being immediately downvoted. I’m not sure if it’s members of the community, or people browsing /all, but I’d ask people to only downvote if the post is unrelated to the community or factually incorrect.
Hi, I’m Fubarberry, and I recently took over moderation of this community due to the original mod being inactive.
This community isn’t very active right now, but I wanted to talk to what members are here about what you want from this community.
For example:
- should this community focus more on AI in general, AI as a tool for creating things (images/video/music/stories/etc), or something else?
- Should this sub be focused on a more positive discussion of AI? Personally I feel like there’s no shortage of places on lemmy to doompost about AI right now, so this community would be more useful discussing beneficial uses of AI. Constructive criticism/acknowledging of issues is fine.
- Should we focus on AI tools available to end-users, and/or discuss AI news in general (AI used for medical research for example).
- Any other thoughts and opinions you have on this community.
Also side note, but right now this community has a bit of an issue with new posts being immediately downvoted. I’m not sure if it’s members of the community, or people browsing /all, but I’d ask people to only downvote if the post is unrelated to the community or factually incorrect.
preview
This is a very small model, ideal for running locally on phones or for small straightforward tasks where speed and efficiency is key. It’s also very quick to fine-tune, you can try fine tuning it yourself in 5 min using a free collab session.
This is a very small model, ideal for running locally on phones or for small straightforward tasks where speed and efficiency is key. It’s also very quick to fine-tune, you can try fine tuning it yourself in 5 min using a free collab session.
preview
It’s not a long article so I suggest reading it, but the basic summary is:
Historically a lot of tech works by compressing data, transferring/storing it, and then uncompressing data.
A lot of LLM use is the opposite, you use an LLM to take your simple statement and write it in a longer more formal format. The person reading it she’s it’s long and wordy, and uses an LLM to summarize it. Instead of compress-decompress, we’re now inflating-deflating.
It’s not a long article so I suggest reading it, but the basic summary is:
Historically a lot of tech works by compressing data, transferring/storing it, and then uncompressing data.
A lot of LLM use is the opposite, you use an LLM to take your simple statement and write it in a longer more formal format. The person reading it she’s it’s long and wordy, and uses an LLM to summarize it. Instead of compress-decompress, we’re now inflating-deflating.
preview
Newelle is a GTK graphical front end for interacting with LLMs from Gnome desktop. It supports both local LLMs and online ones.
In addition to basic chat, it supports:
- Speech to text and TTS models
- Can perform web searches
- Can drag files into Newelle to discuss it’s contents
- Run terminal commands
- Manage files
- Other features like image generation can be added through it’s plugin system.
Obviously you’ll want to be careful with letting it interact with your system directly through files/terminal commands, but still sounds pretty promising.
Newelle is a GTK graphical front end for interacting with LLMs from Gnome desktop. It supports both local LLMs and online ones.
In addition to basic chat, it supports:
- Speech to text and TTS models
- Can perform web searches
- Can drag files into Newelle to discuss it’s contents
- Run terminal commands
- Manage files
- Other features like image generation can be added through it’s plugin system.
Obviously you’ll want to be careful with letting it interact with your system directly through files/terminal commands, but still sounds pretty promising.
preview
The title is slightly misleading, it’s less that all AI models are the same now, and more that they’re converging. The more we improve models, the less variation there is.
The title is slightly misleading, it’s less that all AI models are the same now, and more that they’re converging. The more we improve models, the less variation there is.