news

jimu news reporter zhou dan

on september 19, the 2024 yunqi conference opened in hangzhou. the conference will last for three days, with three main forums and more than 400 sub-forums. at the same time, an exhibition area of ​​about 40,000 square meters will be open, inviting technology enthusiasts from all over the world to enter the smart zone of "cloud-based intelligence and industrial transformation".

wu yongming, ceo of alibaba group and chairman and ceo of alibaba cloud intelligence group, said that in the past summer, alibaba cloud fully supported the paris olympics and achieved a historic breakthrough. for the first time, cloud computing surpassed satellites and became the main broadcasting method for the olympics. ai was also widely used in the olympics for the first time. today, the focus of the yunqi conference is also ai.

the yunqi conference originated in 2009 and was originally named the first china website development forum. it was officially renamed the "yunqi conference" in 2015. jimu news reporters visited the site and found that this year's yunqi conference had more highlights and a stronger sense of technology.

enter keywords and generate hd videos in minutes

this year's yunqi conference opened three theme pavilions: "artificial intelligence + pavilion", "computing pavilion" and "frontier application pavilion".

in the "artificial intelligence + pavilion", the reporter saw that the tongyi exhibition area occupied the c position of the pavilion.

a 13-year-old middle school student typed a few sentences on the tongyi lingma, and a countdown web page written in python was generated in 2 minutes. there were also participants at the scene who used the tongyi lingma to modify the open source contra game code. in 9 code files with more than 2,000 lines of code, they accurately modified the game character's health, jump height and other parameters in just a few minutes.

according to the on-site staff, it usually takes a day for a human programmer to manually develop a web page, including tasks such as requirement analysis, code writing, and testing. now, people only need to input requirements, and tongyi lingma can complete the whole process in 5 minutes. this also means that in the future, as long as you have creativity, you can develop applications and software even if you don't know code.

at the 2024 hangzhou yunqi conference, with the comprehensive improvement of tongyi's big model capabilities, alibaba cloud tongyi lingma ushered in a major upgrade. from only being able to complete basic auxiliary programming tasks a year ago, it has evolved to being able to complete development tasks such as demand understanding, task decomposition, code writing, bug modification, and testing in just a few sentences. it can complete application development from 0 to 1 in as fast as a few minutes, improving development efficiency by dozens of times.

alibaba cloud cto zhou jingren announced on the spot that tongyi wanxiang has been fully upgraded and released a new video generation model that can generate film-quality high-definition videos, which can be applied to film and television creation, animation design, advertising design and other fields. from now on, all users can experience it for free through the tongyi app and tongyi wanxiang official website.

on site, alibaba cloud demonstrated the model's powerful motion generation and concept combination capabilities. by inputting "a cute scene of a little rabbit wearing skates moving flexibly on the ice", tongyi wanxiang was able to generate a high-definition, realistic video in just a few minutes.

the prices of 3 tongyi qianwen main models are further reduced

the reporter also learned on the spot that after the first substantial price cut in may, the three tongyi qianwen main models on the alibaba cloud bailian platform have been reduced in price again. the price of qwen-turbo has dropped by 85% to as low as 0.3 yuan per million tokens, and the prices of qwen-plus and qwen-max have been reduced by 80% and 50% respectively. among them, qwen-plus's reasoning ability is on par with gpt4 and can be applied to complex tasks. it is the best choice for balancing effect, speed and cost. after the price cut, qwen-plus has the highest cost-effectiveness in the industry, and is 84% ​​lower than the industry price of the same scale.

at the same time, alibaba cloud bailian platform also provides more than 50 million tokens and 4,500 image generation quotas for free to all new users.

since the first announcement of the price reduction, the number of paying customers of alibaba cloud bailian has increased by more than 200% compared to the previous quarter. a large number of enterprises and developers have abandoned private deployment and chosen to call various large ai models directly on bailian.

at present, alibaba cloud bailian has served more than 300,000 customers, including many well-known companies and institutions such as faw, kingsoft, liepin, hello group, and the national astronomical observatory. this second round of price cuts by tongyi will continue to increase the penetration rate of ai in all walks of life.

many companies bring the latest technology applications

in the computing pavilion, one booth was extremely popular. many visitors were trying out the recently popular game "black myth: wukong" on the shadowless cloud pc.

"before, if you wanted to play a large-scale game or produce and render a large-scale game, the computer configuration requirements were very high, and you had to buy a very expensive gaming laptop. in fact, what we really buy when we buy equipment is the computing power behind the equipment." the staff of the exhibition area introduced that wuying has created the wuying rubik's cube, which is connected to a unified computing power resource pool, thus skipping the step of purchasing equipment and directly purchasing the computing power behind the equipment.

"our entire shadowless terminal, both the hardware and software client, has been adapted. ipad, android phones, and even the latest vision pro can be adapted." the staff member said that shadowless is mainly used in some daily office and study, and now it has also launched an e-sports mode, which can be applied to the recently popular 3a games such as "black myth: wukong". users do not need to buy a high-configuration computer again to get a smooth gaming experience.

in the frontier applications pavilion, robots from multiple companies are lined up to showcase their latest applications.

the humanoid robot "xiaohui" exhibited by shanghai qingbao engine robot co., ltd. allows the audience to just shout: "xiaohui, xiaohui, i want a cup of latte." "xiaohui" will then quickly make a cup of hot latte and bring it to the audience.

"currently, some coffee shops are using our robot." zheng binbin, a staff member of the exhibition area, introduced that this coffee robot has been put into use in coffee shops in baoshan district, pudong district and other places in shanghai.

in the frontier application pavilion, galaxy general exhibited its first-generation robot, gaibot, which has top-notch generalized object grasping technology, especially in handling transparent and high-gloss objects, with a success rate of over 95%; weijing intelligent exhibited a humanoid picking robot with autonomous path planning and obstacle avoidance capabilities, combined with binocular stereo vision technology, which can achieve efficient and accurate vegetable and fruit picking; star era's robot "xiaoxing max" has 54 degrees of freedom, can walk stably on a variety of complex road surfaces, and perform complex movements. it is the world's first humanoid robot to climb the great wall... full of technology.

the "world's first ai car" p7+ is unveiled

at this year's yunqi conference, "autonomous driving" was undoubtedly one of the hottest topics. not only did tesla, xpeng, leapmotor and other leading domestic and foreign automakers "group together" to participate in the exhibition, but the second topic of the main forum on the first day was also locked on "generative ai reshaping autonomous driving."

all this is happening against the backdrop of the fact that big model technology is being adopted in vehicles at an accelerated pace, and autonomous driving has become the most visible and tangible big model application area for ordinary people.

because of this, liu xiangming, the host of this topic and co-founder of titanium media, threw the first question to the two guests - he xiaopeng, chairman and ceo of xpeng motors, and wu xinzhou, global vice president and head of the automotive business unit of nvidia: "does the big model really improve autonomous driving so much?"

he xiaopeng gave a positive answer to this question. he even listed his feelings at the xiaopeng test track: "with the support of the large model, the autonomous driving test car's steering and cutting are even smoother than my own driving, which is amazing." he xiaopeng even gave a prediction: within 36 months, autonomous driving will be able to drive like an experienced driver.

the upgrade of autonomous driving caused by large models will undoubtedly make the competition among car companies more intense.

he xiaopeng said that a gasoline car has about 30,000 parts, of which about 500 are chips; while the total number of parts of an electric car has dropped to 20,000, but the number of chips is about 5,500. "i think that not many car companies will be able to stay in the game in the future. at present, it is only the initial stage of the competition in autonomous driving. in the future, there will definitely be more breakthroughs."

on the same day, he xiaopeng also drove the "world's first ai car" p7+ to the 2024 yunqi conference. this car is equipped with the industry's leading end-to-end big model. in the past two years, the scale of ai computing power jointly built by xiaopeng motors and alibaba cloud has increased by more than 4 times. he xiaopeng said that he will continue to deepen the ai ​​computing power cooperation with alibaba cloud and accelerate the promotion of end-to-end big models to expand the upper limit of autonomous driving and improve the lower limit.

(source: jimu news)

for more exciting information, please download the "jimu news" client from the application market. please do not reprint without authorization. you are welcome to provide news clues, and you will be paid once adopted.