llms

The DAIR Institute makes sceptical videos warning about the dangerous hype and irresponsible practices currently driving AI, LLMs and related tech. You can follow at:

➡️ @dair

There are already over 70 videos uploaded. If these haven't federated to your server yet, you can browse them all at https://peertube.dair-institute.org/a/dair/videos

You can also follow DAIR's general social media account at @DAIR@dair-community.social

#FeaturedPeerTube #DAIR #AI #LLM #LLMs #OpenAI #SamAltman #Sceptic #Skeptic #PeerTube #PeerTubers

我用来测试 #LLMs #数学 能力的题目已经 #饱和 了。

What if "42" is just the hallucination of an LLM from the future? 🤔

#LLM #LLMs #AI #generativeAI #42

我对传说中的 GPT-5 和 Gemini 2.0 等新 #LLMs 不抱太大希望。大模型在很长一段时间内不会有大的突破，而是 #渐进式 的缓慢优化，不改名用户都感觉不到。这种对数曲线式的进展在几乎所有技术发展中都出现过，只是在大模型这里，第一个突破期显得过快。

At the heart of our discussions about #LLMs is that we have such a poor definition of "intelligence". Too many in the tech space think of it as a simple collection of facts: ask a question, get an answer. If an LLM is really fancy, it breaks the problem up, asks different sub-experts, and collates the replies. It's information retrieval all the way down.

Update2: https://fedihum.org/@lavaeolus/113860035702353891
Update: https://fedihum.org/@lavaeolus/113856096099247855
---
An '#AI-emulation' of Anne Frank made for use in schools.

Who the fuck thought this is appropriate?
Who in the everloving fuck coded this? Who approved it?
Who didn't stop them?

This needs to be luddited 🔥🔥🔥
Spoken as a (digital) historian, who uses #LLMs as tools.

I'm not one quick to anger, but I'm fuming 😤🤬🤬🤬
(Those chatbots are not new, but I hadn't seen this one until this morning in a post by @ct_bergstrom)
#AnneFrank

ALT

ALT

ALT

ALT

为什么 #LLMs 如此强大了实际落地的却还是不多？
可能是因为大部分智力和算力资源用在了模型本身的研究，而应用研究相对不足，就像Altman说的，最大的问题是产品。
下一层原因是模型进步很快，应用跟不上——新模型可以轻松替代你开发的应用层。
从这个意义上说，近来流行的模型撞墙论未必不是一件好事。

Resistance to the coup is the defense of the human against the digital and the democratic against the oligarchic.