Generative AI Is Still Just a Prediction Machine
(hbr.org)
Artificial intelligence tools can now write, code, draw, summarize, and brainstorm. The proliferation of generative AI tools poses serious questions for managers, such as: What tasks can be done by AI, what will humans still need to do, and what are the sustainable sources of competitive advantage as AI continues to improve? To understand the strategic implications of these new capabilities, managers need a framework for when AI will be helpful and when it might fail.
Artificial intelligence tools can now write, code, draw, summarize, and brainstorm. The proliferation of generative AI tools poses serious questions for managers, such as: What tasks can be done by AI, what will humans still need to do, and what are the sustainable sources of competitive advantage as AI continues to improve? To understand the strategic implications of these new capabilities, managers need a framework for when AI will be helpful and when it might fail.
How Did You Do on the AI Art Turing Test?
(astralcodexten.com)
Last month, I challenged 11,000 people to classify fifty pictures as either human art or AI-generated images.
Last month, I challenged 11,000 people to classify fifty pictures as either human art or AI-generated images.
Bringing Red Hat Enterprise Linux to Windows Subsystem for Linux
(redhat.com)
The hybrid cloud is an innovation driver, whether pushing the enterprise technology envelope with breakthroughs like generative AI (gen AI) or simply making traditional IT more efficient and responsive through application modernization.
The hybrid cloud is an innovation driver, whether pushing the enterprise technology envelope with breakthroughs like generative AI (gen AI) or simply making traditional IT more efficient and responsive through application modernization.
Microsoft starts boiling the Copilot frog: It's not a soup you want to drink
(theregister.com)
You think employee-generated content is bad now?
You think employee-generated content is bad now?
Extending the context length to 1M tokens
(qwenlm.github.io)
After the release of Qwen2.5, we heard the community’s demand for processing longer contexts.
After the release of Qwen2.5, we heard the community’s demand for processing longer contexts.
Mistral updates Le Chat with new features
(mistral.ai)
Search, vision, ideation, coding… all yours for free.
Search, vision, ideation, coding… all yours for free.
Pixtral Large
(mistral.ai)
Today we announce Pixtral Large, a 124B open-weights multimodal model built on top of Mistral Large 2. Pixtral Large is the second model in our multimodal family and demonstrates frontier-level image understanding.
Today we announce Pixtral Large, a 124B open-weights multimodal model built on top of Mistral Large 2. Pixtral Large is the second model in our multimodal family and demonstrates frontier-level image understanding.
Google Gemini tells grad student to 'please die' while helping with his homework
(theregister.com)
When you're trying to get homework help from an AI model like Google Gemini, the last thing you'd expect is for it to call you "a stain on the universe" that should "please die," yet here we are, assuming the conversation published online this week is accurate.
When you're trying to get homework help from an AI model like Google Gemini, the last thing you'd expect is for it to call you "a stain on the universe" that should "please die," yet here we are, assuming the conversation published online this week is accurate.
GaussianAnything: Interactive Point Cloud Latent Diffusion for 3D Generation
(nirvanalan.github.io)
GaussianAnything generates high-quality and editable surfel Gaussians through a cascaded 3D diffusion pipeline, given single-view images or texts as the conditions.
GaussianAnything generates high-quality and editable surfel Gaussians through a cascaded 3D diffusion pipeline, given single-view images or texts as the conditions.
Ask HN: Is there any license that is designed to exclude LLMs?
(ycombinator.com)
I don't want my content to be harvested by LLMs; They are removing attribution, among other things. Otherwise, I'd like to stick as close as possible to the open source licenses (say MIT). Is there such a license out there? If not, anyone working on such a thing?
I don't want my content to be harvested by LLMs; They are removing attribution, among other things. Otherwise, I'd like to stick as close as possible to the open source licenses (say MIT). Is there such a license out there? If not, anyone working on such a thing?
Awesome-Geo
(github.com/DavidHuji)
Awesome list for research on GEO (Generative Engine Optimization).
Awesome list for research on GEO (Generative Engine Optimization).
AI-generated poetry is indistinguishable from human-written and more favorably
(nature.com)
As AI-generated text continues to evolve, distinguishing it from human-authored content has become increasingly difficult.
As AI-generated text continues to evolve, distinguishing it from human-authored content has become increasingly difficult.
Show HN: AnyModal – Train Your Own Multimodal LLMs
(github.com/ritabratamaiti)
AnyModal is a modular and extensible framework for integrating diverse input modalities (e.g., images, audio) into large language models (LLMs). It enables seamless tokenization, encoding, and language generation using pre-trained models for various modalities.
AnyModal is a modular and extensible framework for integrating diverse input modalities (e.g., images, audio) into large language models (LLMs). It enables seamless tokenization, encoding, and language generation using pre-trained models for various modalities.
Nvidia presents Llama-Mesh: Generating 3D Mesh with Llama 3.1 8B
(nvidia.com)
This work explores expanding the capabilities of large language models (LLMs) pretrained on text to generate 3D meshes within a unified model.
This work explores expanding the capabilities of large language models (LLMs) pretrained on text to generate 3D meshes within a unified model.
SmoothLLM: Defending Large Language Models Against Jailbreaking Attacks
(arxiv.org)
Despite efforts to align large language models (LLMs) with human intentions, widely-used LLMs such as GPT, Llama, and Claude are susceptible to jailbreaking attacks, wherein an adversary fools a targeted LLM into generating objectionable content.
Despite efforts to align large language models (LLMs) with human intentions, widely-used LLMs such as GPT, Llama, and Claude are susceptible to jailbreaking attacks, wherein an adversary fools a targeted LLM into generating objectionable content.
OpenAI Describes Artists' Use of Glaze and Nightshade as "Abuse"
(80.lv)
While it isn't much, the mere fact that a multi-billion corporation is viewing Artists' use of Glaze and Nightshade as "abuse" speaks volumes, suggesting that the tools do indeed work.
While it isn't much, the mere fact that a multi-billion corporation is viewing Artists' use of Glaze and Nightshade as "abuse" speaks volumes, suggesting that the tools do indeed work.
Google Gemini tells grad student to 'please die' while helping with his homework
(theregister.com)
When you're trying to get homework help from an AI model like Google Gemini, the last thing you'd expect is for it to call you "a stain on the universe" that should "please die," yet here we are, assuming the conversation published online this week is accurate.
When you're trying to get homework help from an AI model like Google Gemini, the last thing you'd expect is for it to call you "a stain on the universe" that should "please die," yet here we are, assuming the conversation published online this week is accurate.
LlamaChunk: Better RAG Chunking Than LlamaIndex
(github.com/ZeroEntropy-AI)
One major pain point of building RAG applications is that it requires a lot of experimentation and tuning, and there are hardly any good benchmarks to evaluate the accuracy of the retrieval step only.
One major pain point of building RAG applications is that it requires a lot of experimentation and tuning, and there are hardly any good benchmarks to evaluate the accuracy of the retrieval step only.
1-Bit AI Infrastructure
(arxiv.org)
Recent advances in 1-bit Large Language Models (LLMs), such as BitNet and BitNet b1.58, present a promising approach to enhancing the efficiency of LLMs in terms of speed and energy consumption.
Recent advances in 1-bit Large Language Models (LLMs), such as BitNet and BitNet b1.58, present a promising approach to enhancing the efficiency of LLMs in terms of speed and energy consumption.
Language agents achieve superhuman synthesis of scientific knowledge
(arxiv.org)
Language models are known to hallucinate incorrect information, and it is unclear if they are sufficiently accurate and reliable for use in scientific research.
Language models are known to hallucinate incorrect information, and it is unclear if they are sufficiently accurate and reliable for use in scientific research.
Something weird is happening with LLMs and chess
(dynomight.substack.com)
A year ago, there was a lot of talk about large language models (LLMs) playing chess. Word was that if you trained a big enough model on enough text, then you could send it a partially played game, ask it to predict the next move, and it would play at the level of an advanced amateur.
A year ago, there was a lot of talk about large language models (LLMs) playing chess. Word was that if you trained a big enough model on enough text, then you could send it a partially played game, ask it to predict the next move, and it would play at the level of an advanced amateur.
Generative AI doesn't have a coherent understanding of the world
(news.mit.edu)
Large language models can do impressive things, like write poetry or generate viable computer programs, even though these models are trained to predict words that come next in a piece of text.
Large language models can do impressive things, like write poetry or generate viable computer programs, even though these models are trained to predict words that come next in a piece of text.
1-bit architecture is turbocharging LLM efficiency
(venturebeat.com)
One-bit large language models (LLMs) have emerged as a promising approach to making generative AI more accessible and affordable.
One-bit large language models (LLMs) have emerged as a promising approach to making generative AI more accessible and affordable.