I know the difference. Neither OpenAI, Google, or Anthropic have admitted they can’t scale up their chat bots. That statement is not true.

That may have been true for the early LLM chatbots but not anymore. ChatGPT for instance, now writes code to answer logical questions. The o1 models have background token usage because each response is actually the result of multiple background LLM responses.

The title of the article is literally a lie which is easily fact checked. Follow the links to quotes in the article to see what the quoted individuals actually said about the topic.

No, a chat bot as it’s talked about here is not an LLM. This article is discussing limitations of LLM training data and inferring that chat bots can not scale as a result. There are many techniques that can be used to continue to improve chat bots.

I’m sorry if I’m coming across as condescending, that’s not my intent. It’s never been “as simple as just throwing more data and CPU at the problem”. There were algorithmic challenges for every LLM evolution. There are still lots of potential improvements using the existing training data. But even if there wasn’t, we’ll still see loads of improvements in chat bots because of other techniques.

Edit: typo

People that don’t understand those terms are using them interchangeably

Yes of course I’m asserting that. While the performance of LLMs may be plateauing, the cost, context window, and efficiency is still getting much better. When you chat with a modern chat bot it’s not just sending your input to an LLM like the first public version of ChatGPT. Nowadays a single chat bot response may require many LLM requests along with other techniques to mitigate the deficiencies of LLMs. Just ask the free version of ChatGPT a question that requires some calculation and you’ll have a better understanding of what’s going on and the direction of the industry.

OpenAI, Google, Anthropic admit they can’t scale up their chatbots any further

Lol, no they didn’t. The quotes this articles are using are talking about LLMs not chatbots. This is yet another stupid article from someone who doesn’t understand the technology. There is a lot of legitimate criticism for the way this technology is being implemented but FFS get the basics right at least.

To be fair, some of my weekends have been a mistake.

I had a really good friend on MySpace that I lost touch with. I think he was a little paranoid, we didn’t speak much and he was always looking over his shoulder. His name was Tom.

This is 100% OPs weird fetish but props to this hussle

I love that whoever filmed this didn’t help or even flinch

The line between webapp and app is very thin these days anyway.

While the user experience may be similar (and in many cases is identical) access to device information is different. For instance, a webapp can not determine the devices volume whereas an Android app can. Device APIs can provide much more confidence that an activity has occurred. I doubt this was an arbitrary decision or gate-keeping by the developers.

Pointless gatekeeping by the devs

More likely it’s to ensure that you’ve actually listened to the podcast you’re rating

edit: Can someone explain the downvotes, what am I missing here?

I find I have much less Internet fights in the fediverse than I used to on commercial “social” media. Reddit and Facebook would bring out the worst of my pettiness 😅 I have had many lovely conversations on here.

That data is available is the ActivityPub but as far as I know, you don’t have access to the vote data as a Lemmy user. You would have access to this data if you were running your own Lemmy server.

But honestly, let it go. Someone stole fake Internet points from you. It costs you nothing but it cost them 5 minutes of their lives. You’re having a fight with some stranger that doesn’t value their own time. Value your time and stop feeding these kinds of energy vampires. You deserve better! Have a wonderful day my friend :)

Did someone get in a petty Internet fight?

That’s unlikely in a closed heat exchange system. Maybe some additional evaporation because the water is slightly warmer. But unless I’m missing something, it seems very misleading to suggest that a Bitcoin transaction uses 16 kilolitres because of evaporation. Napkin math, it would require about 10 megawatt/hours of energy to evaporate that much water (please correct me if I’m wrong). I’m not a Bitcoin fanboy, I just don’t like BS.

These data centers consume water for cooling systems

How does a data center consume water? Doesn’t every liter that enters as freshwater leave as slightly warmer freshwater? What am I missing here?

I used to get shin splints so now I only jog on jello