March 12 news, according to foreign media TechCrunch reported today, IBM CEO Arvind Krishna said he thinks AI has important value, but not everything. Anthropic CEO Dario Amodei predicted that "in the next three to six months 90% code will be generated by AI", Krishna does not agree. "I think it's more likely that the percentage of code generated by AI will be 20% to 30%, which is much less than 90%," Krishna said. "Some simple tasks can be done by A...

Just last night, the news that Anthropic was releasing a new model started to spread widely in the AI community, but not the expected Claude 4.0, but the 3.7 Sonnet version. Early this morning, Anthropic's new flagship model arrived as promised, officially releasing its smartest model to date and the first hybrid inference model on the market -- Claude 3.7 Sonnet. After learning about it, there are roughly two things: ① Hybrid inference model 3.7sonnet ② claude code code work ...

February 24, 2011 - While OpenAI CEO Sam Altman insists that AI models will be able to outperform "low-level" software engineers by the end of the year, a new study by the company's researchers suggests that even the most advanced AI models can't compete with human programmers. human programmers. In a new paper, the researchers note that even cutting-edge models -- the most innovative and groundbreaking AI systems -- "still cannot solve most" programming tasks. To that end, the researchers developed a program called...

On January 4, Ali Tongyi Qianqian Qwen newly launched the CodeElo benchmark test to assess the programming level of Large Language Model (LLM) by comparing the Elo rating system with human programmers. Project Background One of the AI scenario applications of Large Language Model is to generate and complement code, but there are many challenges in assessing the real ability of programming at this stage. Existing benchmarks including LiveCodeBench and USACO have limitations, lack robust private test cases, do not support specialized judgment systems,...

Devin, the world's first "AI programmer," opened its doors on December 11th, less than a year after its debut. Its developer, Cognition Labs, announced that individual developers and engineering teams can get full access for $500 per month (currently around RMB 3,635), with a subscription that includes Slack integration, IDE extensions, and specialized technical support. For large enterprises, the company offers a customized Enterprise Edition solution. Devin is fluent in multiple programming languages and works independently on websites and applications...

The rapid development of artificial intelligence (AI) in recent years has sparked widespread discussion and concern about whether AI will replace human jobs. Artists and programmers are particularly worried, as AI seems to be capable of doing both jobs. However, Sundar Pichai, CEO of Google and its parent company Alphabet, believes that AI will not replace programmers, but instead will help more people become programmers. In a speech at Carnegie Mellon University, Pichai talked about the impact of AI on programming jobs, saying that AI is most likely to help people,...

Over the past few years, computer programmers have turned to chatbots like OpenAI's ChatGPT to help them code, dealing a blow to sites like Stack Overflow and causing it to have to lay off nearly 301 TP3T employees last year. The problem, however, is that a team of researchers from Purdue University released their findings at the Computer-Human Interaction Conference and found that 521 TP3T of the answers to programming questions generated by ChatGPT were wrong. Image source note: Image generated by AI, image licensed to service provider Midjourne...

Programmer Q&A platform Stack Overflow and OpenAI today announced a new API partnership. OpenAI and Stack Overflow will provide OpenAI users with the accurate, audited data foundation they need through the Overflow API so that AI tools can quickly find solutions to problems, allowing technicians to focus on high-priority tasks; OpenAI will also display questions from Stack Overflow directly in ChatGPT…

Alibaba Cloud is promoting AI programming internally, using Tongyi Lingma to assist programmers in writing code, reading code, checking bugs, optimizing code, etc. Alibaba Cloud has also assigned a formal employee number to Tongyi Lingma - AI001. According to a person from Alibaba Cloud, "The company's future 20% code will be written by Tongyi Lingma, but programmers will still be the core of R&D, and they will have more time to focus on system design and core business development." It is reported that under the traditional development model, programmers need to spend a lot of energy every day writing repetitive code, debugging and optimizing, writing code comments, etc.

Stability AI recently launched the Stable Code Instruct 3B large language model for programmers. Its highlight is that it can translate from one programming language to another. The functions of the Stable Code Instruct 3B large language model are as follows: The model can automatically generate code based on the code context. Based on the relevant code context, it automatically inserts and completes the relevant code. Stable Code Instruct 3B can generate code that interacts with the database…

On March 9, on CCTV's "Dialogue" program, Baidu founder, chairman and CEO Robin Li said that there will be no such profession as "programmer" in the future, because as long as you can speak, everyone can have the ability of a programmer. "In the future, there will only be two programming languages, one is called English and the other is called Chinese." Robin Li said. Speaking of the "100 Model Wars", Robin Li believes that this is a huge waste of social resources. Don't "roll" big models, go "roll" applications, only applications can truly create value directly. At the same time, he emphasized that when making applications used by hundreds of millions of users...

NVIDIA, in conjunction with Hugging Face and ServiceNow, recently announced a family of LLMs called StarCoder2, which it hopes will become the new standard in code generation, offering a number of advantages, including performance, transparency and cost-effectiveness. The family of models includes a 3 billion parameter model trained by ServiceNow, a 7 billion parameter model trained by Hugging Face and a 15 billion parameter model trained by NVIDIA. This is accomplished through the use of an algorithm called Stack v...