


What will the AI industry evolve into in 2024?
OpenAI Chairman Greg Brockman once predicted on the last day of last year: 2023 will make 2022 look like AI development has not yet awakened.
Sure enough, in 2023, the AI industry will usher in a full-scale explosion.
Yesterday, Clement Delangue, CEO of Hugging Face, the world’s largest AI open source community, made 6 specific predictions for the development of the industry in 2024:
1. A certain popular AI company will go bankrupt or be acquired at a very low price.
Open source LLM has a level of capability comparable to closed source LLM
3.AI in video, biology, chemistry, time series and other fields will bring about huge breakthroughs.
4. The public will be more concerned about the economic and environmental costs of AI.
5. In the future, mass media will be filled with content generated by artificial intelligence
6. 10 million AI developers on Hugging Face will It will create new job opportunities and will not lead to a surge in unemployment.
The breakthroughs in the AI industry in 2023 are mainly reflected in the AI technology itself. However, these six new predictions indicate that AI technology will further break through the circle in 2024, and its breakthrough influence will far exceed the scope of the AI industry.
Netizens responded to his six predictions The predictions were evaluated and the probability of occurrence of three of them was considered to be no more than 50%
Some people think that all six predictions are credible, even Several of them have become reality
The first wave of AI companies will go bankrupt
For the first prediction, combined with the previous two days OpenAI's violent turmoil has caused netizens to become angry.
Clement quickly came out to smooth things over. I was predicting myself.
Netizens have made various speculations, mentioning star startups such as Adept and Perplexity
However, in fact, similar The situation has already occurred as early as 2023. Jasper AI is an AI startup that was once valued at more than $1.5 billion. As the "originator of shelling GPT", the company reported negative news about layoffs and an 80% drop in valuation in July
In OpenAI After the launch of GPTs, it can be expected that the prospects of various artificial intelligence companies based on OpenAI technology will become increasingly limited
If they cannot find a path to independently create value in the future, From a valuation of more than 1 billion US dollars to bankruptcy or being acquired at a low price is not too surprising
Open Source VS. Closed Source
The gap between AI open source and closed source, the future Whether it will be further expanded or reduced, industry leaders, AI open source companies, scientific researchers, and users have always had different voices. People have always had different views on whether the gap between AI open source and closed source will expand or shrink in the future. Industry leaders, AI open source companies, researchers and users have always had different views.
From Google engineers at the beginning of the year revealed that neither OpenAI nor Google has a moat, and open source AI is their biggest opponent.
On the other hand, open source models that continue to emerge in various fields claim to be close to or exceed GPT-4
Recently, the Berkeley team announced the Starling-7B model, which surpasses all other models in some benchmark tests and is close to the level of GPT-4 by using RLAIF technology
Project address: https://starling.cs.berkeley.edu/
There are even some open source models that claim to be comparable to GPT-4 on some specific tasks, including even a 7B size model
Supporters of the open source model claim that compared with adopting a closed source model, the difference in the open source model may take 3-5 years to be reflected
The second prediction caused netizens to Their controversy
"Due to the huge gap in computing power between open source and closed source, it is still very difficult for open source to catch up with closed source."
"I don’t understand why the gap between open source and closed source will shrink. After all, all the knowledge and technology of the open source model are shared, while closed source AI always has something unique to them. 》
AI For Science
In the biological field, DeepMind’s AlphaFold is already surpassing human capabilities At a high level, the structure of proteins was predicted, directly breaking through the forefront of biology.
Microsoft recently published a report, trying GPT-4 in the fields of biology, computational chemistry, drug discovery, materials design and The scientific research fields of partial differential equations (PDE) are valuable as academic research assistants.
Paper link: https://arxiv.org/abs/2311.07361
According to Microsoft According to researchers, artificial intelligence tools will greatly accelerate the progress of basic scientific research
Currently, there are many tools and platforms that apply machine learning technology to various professional fields
Project address: https://chemintelligence.com/our-platform
Netizens are also concerned about the breakthroughs that AI may achieve in scientific research fields related to time series (Time-Series, which refers to the use of time series in applied science and engineering fields such as statistics, signal processing, pattern recognition, econometrics, and mathematical finance). Very optimistic.
Some netizens even contributed a paper to explain why time series related scientific research is so important
The content that needs to be rewritten is: Paper link: https://arxiv.org/abs/2310.03589
"Finally, someone has seen this, and a good Transformer model for time series has emerged."
The energy caused by AI And environmental costs
#In a podcast, Musk talked about the most important first-principles thinking in the AI industry. He believes that there is still a lot of room for improvement in the Transformer model in terms of energy and intelligence output ratio
Research by Alex de Vries, a data scientist at Vrije Universiteit Amsterdam in the Netherlands, predicts that by 2027 , the artificial intelligence server farm can use 85 to 134 terawatt hours of energy per year
https://www.php.cn/link/f52166cd701447355be87cbf41d31ca4
And foreign media also have Many reports have pointed out that the development of AI technology will cause a sharp rise in energy and water consumption in the future.
In 2027, AI energy consumption may be equivalent to the annual electricity consumption of Argentina, the Netherlands or Sweden, or 0.5% of global energy demand.
AIGC floods the media
It is no exaggeration to say that AI-generated content of videos and images, Now moving at the speed of light.
One to two years ago, it was difficult for anyone to imagine that such an AI animation could be completely generated by AI.
The co-founder of DreamWorks recently publicly stated that artificial intelligence will reduce the cost of the animation industry by 90% in the next three years!
Maybe next year, animation content creation will be further democratized, and the consequences will be that, like short videos overnight Taking the world by storm, videos generated by Gen AI will become a very important part of video content.
AI’s impact on the labor market
The impact of AI on the labor market has always been a concern for the AI industry. The focus of controversy.
To some extent, the Hugging Face CEO’s prediction may be a potential answer to this question.
If more people can join the open source AI developer industry, will the jobs created thereby be able to make up for the labor force replaced by the development of AI technology.
But most netizens don’t seem to agree with this logic.
The second and sixth items are both jokes. Although more developers will appear on Hugging Face, AI will still reconstruct the labor market.
After all, Sam Altman was briefly unemployed for 2 days after becoming the CEO of OpenAI because of AI.
The above is the detailed content of Prediction: Hugging Face CEO reveals six major changes in the AI industry in 2024!. For more information, please follow other related articles on the PHP Chinese website!

1 前言在发布DALL·E的15个月后,OpenAI在今年春天带了续作DALL·E 2,以其更加惊艳的效果和丰富的可玩性迅速占领了各大AI社区的头条。近年来,随着生成对抗网络(GAN)、变分自编码器(VAE)、扩散模型(Diffusion models)的出现,深度学习已向世人展现其强大的图像生成能力;加上GPT-3、BERT等NLP模型的成功,人类正逐步打破文本和图像的信息界限。在DALL·E 2中,只需输入简单的文本(prompt),它就可以生成多张1024*1024的高清图像。这些图像甚至

Wav2vec 2.0 [1],HuBERT [2] 和 WavLM [3] 等语音预训练模型,通过在多达上万小时的无标注语音数据(如 Libri-light )上的自监督学习,显著提升了自动语音识别(Automatic Speech Recognition, ASR),语音合成(Text-to-speech, TTS)和语音转换(Voice Conversation,VC)等语音下游任务的性能。然而这些模型都没有公开的中文版本,不便于应用在中文语音研究场景。 WenetSpeech [4] 是

“Making large models smaller”这是很多语言模型研究人员的学术追求,针对大模型昂贵的环境和训练成本,陈丹琦在智源大会青源学术年会上做了题为“Making large models smaller”的特邀报告。报告中重点提及了基于记忆增强的TRIME算法和基于粗细粒度联合剪枝和逐层蒸馏的CofiPruning算法。前者能够在不改变模型结构的基础上兼顾语言模型困惑度和检索速度方面的优势;而后者可以在保证下游任务准确度的同时实现更快的处理速度,具有更小的模型结构。陈丹琦 普

由于复杂的注意力机制和模型设计,大多数现有的视觉 Transformer(ViT)在现实的工业部署场景中不能像卷积神经网络(CNN)那样高效地执行。这就带来了一个问题:视觉神经网络能否像 CNN 一样快速推断并像 ViT 一样强大?近期一些工作试图设计 CNN-Transformer 混合架构来解决这个问题,但这些工作的整体性能远不能令人满意。基于此,来自字节跳动的研究者提出了一种能在现实工业场景中有效部署的下一代视觉 Transformer——Next-ViT。从延迟 / 准确性权衡的角度看,

3月27号,Stability AI的创始人兼首席执行官Emad Mostaque在一条推文中宣布,Stable Diffusion XL 现已可用于公开测试。以下是一些事项:“XL”不是这个新的AI模型的官方名称。一旦发布稳定性AI公司的官方公告,名称将会更改。与先前版本相比,图像质量有所提高与先前版本相比,图像生成速度大大加快。示例图像让我们看看新旧AI模型在结果上的差异。Prompt: Luxury sports car with aerodynamic curves, shot in a

人工智能就是一个「拼财力」的行业,如果没有高性能计算设备,别说开发基础模型,就连微调模型都做不到。但如果只靠拼硬件,单靠当前计算性能的发展速度,迟早有一天无法满足日益膨胀的需求,所以还需要配套的软件来协调统筹计算能力,这时候就需要用到「智能计算」技术。最近,来自之江实验室、中国工程院、国防科技大学、浙江大学等多达十二个国内外研究机构共同发表了一篇论文,首次对智能计算领域进行了全面的调研,涵盖了理论基础、智能与计算的技术融合、重要应用、挑战和未来前景。论文链接:https://spj.scien

译者 | 李睿审校 | 孙淑娟近年来, Transformer 机器学习模型已经成为深度学习和深度神经网络技术进步的主要亮点之一。它主要用于自然语言处理中的高级应用。谷歌正在使用它来增强其搜索引擎结果。OpenAI 使用 Transformer 创建了著名的 GPT-2和 GPT-3模型。自从2017年首次亮相以来,Transformer 架构不断发展并扩展到多种不同的变体,从语言任务扩展到其他领域。它们已被用于时间序列预测。它们是 DeepMind 的蛋白质结构预测模型 AlphaFold

说起2010年南非世界杯的最大网红,一定非「章鱼保罗」莫属!这只位于德国海洋生物中心的神奇章鱼,不仅成功预测了德国队全部七场比赛的结果,还顺利地选出了最终的总冠军西班牙队。不幸的是,保罗已经永远地离开了我们,但它的「遗产」却在人们预测足球比赛结果的尝试中持续存在。在艾伦图灵研究所(The Alan Turing Institute),随着2022年卡塔尔世界杯的持续进行,三位研究员Nick Barlow、Jack Roberts和Ryan Chan决定用一种AI算法预测今年的冠军归属。预测模型图


Hot AI Tools

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Undress AI Tool
Undress images for free

Clothoff.io
AI clothes remover

AI Hentai Generator
Generate AI Hentai for free.

Hot Article

Hot Tools

Dreamweaver CS6
Visual web development tools

SecLists
SecLists is the ultimate security tester's companion. It is a collection of various types of lists that are frequently used during security assessments, all in one place. SecLists helps make security testing more efficient and productive by conveniently providing all the lists a security tester might need. List types include usernames, passwords, URLs, fuzzing payloads, sensitive data patterns, web shells, and more. The tester can simply pull this repository onto a new test machine and he will have access to every type of list he needs.

MantisBT
Mantis is an easy-to-deploy web-based defect tracking tool designed to aid in product defect tracking. It requires PHP, MySQL and a web server. Check out our demo and hosting services.

mPDF
mPDF is a PHP library that can generate PDF files from UTF-8 encoded HTML. The original author, Ian Back, wrote mPDF to output PDF files "on the fly" from his website and handle different languages. It is slower than original scripts like HTML2FPDF and produces larger files when using Unicode fonts, but supports CSS styles etc. and has a lot of enhancements. Supports almost all languages, including RTL (Arabic and Hebrew) and CJK (Chinese, Japanese and Korean). Supports nested block-level elements (such as P, DIV),

ZendStudio 13.5.1 Mac
Powerful PHP integrated development environment
