Strategy information_Software game information strategy collection

English

中文(简体) 中文(繁体) 한국어 日本語 English Português Español Русский العربية Indonesia Deutsch Français ภาษาไทย

Home> Strategy information

All Software strategy Mobile game strategy Single-player strategy Online game strategy Arcade strategy Esports

Strategy information

Google releases video generation model Lumiere, with excellent motion range and consistency performance

Google released the video generation model Lumiere, which uses the innovative U-Net architecture to complete video generation in one go and takes the lead in text-to-video conversion. The model shows excellent performance in various video editing tasks. F

2025-01-30
Microsoft GenAI team focuses on building smaller, cheaper AI models

The Microsoft GenAI team is working on building smaller, more economical language models to reduce the cost and improve scalability of running artificial intelligence models on supercomputers. The team will be led by Misha Bilenko and will become part of

2025-01-30
Anthropic fights back against music publisher injunction motion: Using copyrighted content as LLM training data is justified

The music publisher filed a preliminary injunction against Anthropic, and Anthropic responded by pointing out that the use of copyrighted content as LLM training data was reasonable and rejecting the injunction request. The company believes that training

2025-01-30
Artisse AI raises $6.7 million to build more realistic AI photography applications

Artisse, an AI photography app, announced that it has successfully raised $6.7 million to focus on improving the realism of its AI-generated photos. Users can upload photos, train the AI model, and use text or image prompts to generate new photos of var

2025-01-30
Meta teamed up with New York University to release the OK-Robot robot framework to realize the task of cleaning up the home environment

Meta collaborated with New York University to release the OK-Robot robot framework, which incorporates cutting-edge visual language models and has a success rate of 82%. OK-Robot does not require pre-training and can run in real home environments, providi

2025-01-30
MIT study: AI won’t replace your job yet because it’s still too expensive

New research conducted by MIT finds that the high cost of artificial intelligence technology may make employers more inclined to retain existing employees. Even if the cost of deploying AI will fall as the technology develops, it will still be decades bef

2025-01-30
Google terminates contract with Appen, an AI data company that helped train products like Bard

Google announced the termination of its contract with Australian data company Appen, which has provided large-scale language model AI training for Google products such as Bard and Search. The decision caused Appen to face a significant revenue impact and

2025-01-30
ComfyUI launches new InstantID function, providing 9 style choices

ComfyUI recently launched the InstantID function to make role maintenance simpler and more efficient. Users can choose from 9 different styles to maintain personalized roles. Supports local and hub models, combined with artistic visualization workflow, pr

2025-01-30
University of Washington launches agent tuning method to improve the efficiency of large model tuning

Webmaster Home reported that the University of Washington launched a proxy tuning method that can achieve efficient tuning of large models without touching the model weights by comparing the prediction results of small adjusted models and unadjusted model

2025-01-30
Which sports have the undisputed goat?

In the sports world, "GOAT" (Greatest Of All Time) is a high praise for athletes who have achieved outstanding achievements. In the field of basketball, Michael Jordan is generally regarded as the GOAT; in the field of football, Messi and Ronald

2025-01-30
Real-time dynamic alignment of values without training: Submit the open source values alignment method OPO, applicable to both closed source and open source large models

With the development of artificial intelligence technology, large language models represented by GPT-4 are having a profound impact on society with their powerful capabilities. The new method OPO does not require retraining the model, dynamically aligning

2025-01-30
Yann LeCun: Generative models are not suitable for processing videos, AI has to make predictions in an abstract space

In this article, Yann LeCun, speaking at the World Economic Forum, said that currently widely used generative models are not suitable for processing video data. Video data is a new challenge for AI. AI understanding of video cannot rely on prediction in p

2025-01-30
How to develop an independent Taoke app

Developing an independent Taoke APP involves many key steps, including market research, identifying target users, selecting technology stacks, designing UI/UX, developing front-end and back-end, functional implementation, security optimization, and pre-la

2025-01-30
"Coaxing Simulator" explodes in popularity, with 700,000 users coaxing their cyber girlfriends online! 1 billion tokens burned in one day

"Coaxing Simulator" is a popular AI application, attracting 700,000 users to coax their cyber girlfriends online within 24 hours. In this game, users need to respond to scenes of couples quarreling and try to coax their virtual girlfriends/boyfr

2025-01-30
Yann LeCun: Generative models are not suitable for processing videos. AI understands videos and makes predictions in an abstract space.

Yann LeCun, Turing Award winner and Meta's chief AI scientist, pointed out at the World Economic Forum that generative models are not suitable for processing videos, and AI needs to make predictions in an abstract space. As text data on the Internet i

2025-01-30

Prev 5 6 7 8910 11 12 13Next