Artificial intelligence (AI) has become the hottest technology field in many countries, with governments and companies rushing to invest heavily in the field, spurring innovation.
In addition, the COVID-19 pandemic has forced us to rely even more on technology, online activity and artificial intelligence. Among them, ARTIFICIAL intelligence is particularly important for enterprises, which can realize personalized services on a large scale and meet customers’ ever-increasing experience needs.
In a report published on March 15, Forbes.com listed the top five ai trends we are looking forward to in 2021. These include the proliferation of low-code/no-code tools and the increasingly “accessible” way in which children can easily create their own AI.
Low code/no code tools
Automatic machine learning (AutoML) is not new. In 2020, Huawei will hire a doctor of machine learning with a salary of 1 million yuan, and one of the research directions will be AutoML.
Machine learning allows algorithms to automatically extract a set of rules from data to extract relevant features. With the development of machine learning, more and more human intervention is required, while AutoML automates the whole process of machine learning models from construction to application.
While AutoML can build high-quality AI models without solid knowledge of data science, the low-code/no-code platform goes a step further — it can build entire production-grade AI-driven applications without deep programming knowledge. Last year saw the emergence of low-code/no-code tools in a variety of applications, from building applications to vertical AI solutions for the enterprise, and this trend is expected to continue this year.
Low code/no code tools are the next frontier for tech giants, a $13.2 billion market that is expected to grow to $45.5 billion by 2025, according to data.
This is exemplified by Amazon’s Honeycode platform, which was launched in June 2020. Honeycode, a code-free development environment similar to a spreadsheet interface, has been called a “boon” for product managers.
Advanced pre-trained language model
Bidirectional encoder Representations from Converters (BERT) is a new language model developed and released by Google at the end of 2018. As a newcomer to the field of natural language processing (NLP), BERT has become the embodiment of NLP’s major advances in the past few years, and he has set a new record in 11 NLP tests, even surpassing human performance.
In recent years, pre-trained language models similar to BERT models (such as question answering, named entity recognition, natural language reasoning, text classification, etc.) have played an important role in many natural language processing tasks.
These pre-trained language models are powerful and revolutionize language translation, understanding, summarization, and so on, but they are expensive and time consuming to train. The good news is that advanced pre-training models can lead to a new generation of ai services that are efficient and easy to build.
Gpt-3 is the best of them all! It is a natural language processing model built by OpenAI with huge investment. With 175 billion large parameters, it is the strongest AI model in NLP field. Gpt-3 has been popular across media platforms since it was first launched in May last year, thanks to its amazing text-generating capabilities. Not only can it answer questions, write essays, write poems and translate articles, but it can also generate code, do mathematical reasoning, analyze data, draw diagrams and make resumes, and even play games — and it works surprisingly well.
Synthetic content generation
Algorithmic innovation in artificial intelligence is not unique to NLP. Generative adversarial networks (GANs) have also seen a spate of innovations, demonstrating the remarkable success of scientists in creating art and fake images.
GANs, first proposed by Ian Goodfellow, an AI scholar at the University of Montreal in Canada, are also complex to train and adapt because they require large data sets to train. But scientists’ innovations dramatically reduced the amount of data needed to create GANs. Nvidia, for example, has demonstrated a new way to increase the efficiency of training GANs, requiring less data than previous methods. This allows GANs to be used in a wide range of fields, from medical applications such as synthesising histological images of cancer to deeper “Deep fakes”.
Deepfakes are a type of high-powered hacking technology that uses the latest artificial intelligence technology to allow ordinary people to create video clips with the help of computers, and the faces in the videos can be transformed into anyone’s face. “The so-called success is nothing, the failure is nothing”, the video “face change” has aroused great attention, but also caused great controversy. Just five days after its launch, the hack was widely hated and then banned worldwide.
Artificial intelligence for children
With the popularity of low-code tools, AI creators tend to be younger. An elementary school student can now create artificial intelligence for his own use — from classifying text to drawing images. American high schools already offer ai courses, and junior high schools are not far behind.
At the 2020 Synopsys science fair in Silicon Valley, for example, 31% of winning software projects used AI in their innovations. Even more impressively, 27 percent of these aI were created by students in grades 6 through 8. One of the winners was an eighth-grader who created a convolutional neural network that can detect diabetic retinopathy from eye scans.
Machine learning operation
Machine learning operations (MLOps) is a relatively new concept in artificial intelligence, involving the best managed data scientists and operators to effectively develop, deploy, and monitor models.
In 2020, huge changes in operational workflow, inventory management, traffic patterns and more caused many AI to have unexpected behavior due to the COVID-19 pandemic, which is known as drift — a mismatch between input data and ai training expectations. While companies deploying machine learning in production have faced many challenges before, such as drift, the COVID-19 pandemic has increased demand for MLOps. Similarly, with the implementation of privacy regulations such as the California Consumer Privacy Act of 2018, governance and risk management are increasingly needed for companies that operate on customer data. According to some data, the market size of MLOps is expected to reach $4 billion by 2025.
These are not all new trends in AI, but they are worth noting because they highlight three important aspects. First, ai is increasingly being used in the real world, as evidenced by the problems caused by COVID-19 and the growth of MLOps. Second, people are constantly innovating in the field, as BERT and GANs did. Finally, the threshold for the creation of ARTIFICIAL intelligence is getting lower and lower, laying a solid foundation for its “flying into the ordinary home”.
The ideal and future of ARTIFICIAL intelligence is always good, but despite the above innovations, we still need to promote and guide its development in a down-to-earth way, so that it can better benefit mankind.
At present, the above content has been packaged into a complete e-book, and the specific access is as follows:
Method of data Collection
Follow the public account [Pegasus Club]
Background reply number [124]
You can view the download method
▼
Past welfare
Pay attention to the pegasus public number, reply to the corresponding keywords package download learning materials; Reply “join the group”, join the Pegasus AI, big data, project manager learning group, and grow together with excellent people!
From beginning to research, the 10 most Readable books in the field of artificial intelligence
AI Artificial Intelligence/Big Data /Database/Linear Algebra/Python/ Machine Learning /Hadoop
Reply number “12” small white | Python + + machine learning Matlab neural network theory + practice + + + depth video + courseware + source code, download attached!
Reply number “14” small white | machine learning and deep learning required books + machine learning field video/PPT + large data analysis books recommend!
Respond to digital “29” dry | 28 this big data/data analysis, data mining ebook collection of free download!
Reply to the number “16” 100G Python from beginner to Master! Complete video tutorials + Python Classics for self-study!
Reply number “36” 286 pages PDF teach you how to understand deep learning algorithms, theories and computing systems! (Downloadable)
Reply digital download “37” timed | 265 g python a full set of video tutorials, from entry into order to interview skills!
Reply number “38” blockbuster | Wu En of new book “the Machine Learning Yearning, 58 latest chapter 1 – (download PDF)
Answer number “46” 700 pages of Graphics Deep Learning (with code and download)
The number “47” reply good books recommend | 106 pages of the Python advanced Chinese version download (attached)