


As soon as next month! GPT-4, the strongest language model on the planet, will be released soon? CEO hints: Passed the Turing test
As the most anticipated AI model in history, Is GPT-4 really coming?
# In the past few days, a column by AI analyst Romero quickly "exploded" the entire AI technology circle as soon as it was released.
He said, "GPT-4 is almost ready and will be released at some point between December this year and February next year".
There are even rumors that GPT-4 has advanced to the point of being "no different from humans", even 「passed the Turing test」. Compared with before, the introduction cost of enterprises will also drop significantly.
Currently, there is no model in the entire AI industry that can "officially and definitely" pass the Turing test. Therefore, the launch of GPT-4 is likely to become another great achievement in the history of AI. One year.
As soon as next month, the fourth generation model is coming soon?
In June 2020, OpenAI, a world-class artificial intelligence research institution founded by Elon Musk and others, released the third-generation generative pre-training model GPT-3.
GPT-3 has 175 billion parameters, which is 100 times that of the previous generation model GPT-2. It also improves the previous parameter record of similar NLP models by 10 times.
The release of GPT-3 is regarded as a major development by the industry, and has therefore been listed as one of the "Top Ten Breakthrough Technologies in 2021" by "MIT Technology Review".
The huge performance leap has also given everyone high hopes for OpenAI’s next-generation generative pre-training model.
If GPT-4 improves GPT-3 as much as GPT-3 does GPT-2, then this will be an improvement close to 100 times.
But in the past two years, OpenAI has always been very "shy" when talking about GPT-4, and it can even be said that it has remained silent most of the time. We can only "get a glimpse" from the sporadic information that management inadvertently revealed.
The last time news about GPT-4 leaked out, it dates back to the beginning of this year.
OpenAI CEO Sam Altman said at the time that GPT-4 was expected to be released between July and August this year.
However, recently, American technology blogger Robert Scoble said:
"Disruption is coming. GPT- 4 is better than anyone expected. It is one of several major AI products due to launch next year."
Cambrian AI analyst Alberto Romero ( Alberto Romero also quoted a tweet from Silicon Valley engineer Igor Baikov in September in a column in "Algorithm Bridge":
" OpenAI is already training GPT-4 and plans to release it between December and February."
In addition, the first-generation GPT, which contains 117 million parameters, was released in 2018; In 2019 and 2020, we ushered in GPT-2 and GPT-3 respectively.
OpenAI, which has been "reserved" for more than a year, may be time to introduce something new.
"I wouldn't be surprised if GPT-4 can pass the Turing test"
Regarding GPT-4, a representative of an AI startup company analyzed that, "GPT -4 easily passed the Turing test, or indirectly indicates that the performance is high enough to have the potential to pass the Turing test."
A recent tweet from OpenAI CEO Altman, This rumor seems to have been confirmed from the side.
He uploaded a photo of Star Wars with a sentence:
"Don't settle for the 'techno-horror' you created" And pride, the ability to pass the Turing test is powerless in front of the "Force"."
This is an imitation of the classic lines of Darth Vader, the famous character in the movie Star Wars series.
Romero also said in the article:
"If GPT-4 can fully pass Turing test, I wouldn't be surprised."
The Turing test is a test designed to understand how naturally computers communicate with humans.
Currently, there is no AI model in the AI community that has "formally and definitely" passed the Turing test.
Although some artificial intelligence models claimed to have passed the Turing test in the past, because its standards and rules were not clear, they used some "blind spots" and did not pass the test overall. test.
If the rumors are true and GPT-4 can pass the Turing test without restrictions, then it will become a well-deserved"First in history 》.
Training costs are significantly reduced and commercialization potential is huge
There is another reason why the AI circle has paid widespread attention to GPT-4: it is said that GPT-4 will be introduced into commercial services in the future The actual threshold will be much lower than the existing GPT-3 model.
At the same time, it is expected that companies that are unable to use AI technology due to the huge cost and required infrastructure will also have more room to use GPT-4.
According to Romero’s prediction, OpenAI will focus its next-generation models more on optimizing data processing rather than parameter size, so the scale of GPT-4 is likely Similar to GPT-3.
So far, advances in artificial intelligence technology and expansion of model sizes have been carried out together, but adding features to lightweight models in the near future is considered promising.
GPT-4’s model training costs are expected to be lower than before. "The training cost of GPT-4 is expected to be significantly lower than that of GPT-3," Romero wrote.
As we all know, GPT-3 costs hundreds of thousands per class just in the learning phase. This is why large companies cannot easily release GPT-3-based artificial intelligence services for ordinary users.
In South Korea, only SK Telecom’s personal artificial intelligence service “Adot” is based on GPT-3.
Romero also said that the cost of each training of GPT-4 is expected to drop below $1 million.
An ICT company official said:
"If all the known information is true, then there will be several types of products next year. Chatbot services. They are so natural that users don’t know whether they are talking to a human or a computer."
"Artificial intelligence services may increase."
Netizens: Need to be wary
Regarding OpenAI’s next-generation “Thanos”-level model, many netizens have expressed that they are “too much looking forward to it.”
However, Gary Marcus, a well-known AI scholar and founder and CEO of Robust.AI, once again badmouthed The future of GPT-4.
He said on social media: "There are rumors that GPT-4 is coming. Here is my prediction: It will still have serious problems of fiction/misinformation, toxicity and bias."
Marcus is well known for his concern for the AI industry, but he has also expressed many opinions such as "deep learning hits a wall" and "GPT-3 is completely meaningless", opposing the hype of AI.
Some netizens also expressed concerns about the potential "power" of the next-generation model:
"If this is true, Then 'his' IQ has reached at least 100."
"This will be one of the greatest technological revolutions in history, but society needs to pay attention to it immediately. Because 'he' 』The applicable potential is unlimited.」
The above is the detailed content of As soon as next month! GPT-4, the strongest language model on the planet, will be released soon? CEO hints: Passed the Turing test. For more information, please follow other related articles on the PHP Chinese website!

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

Video Face Swap
Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Hot Tools

Notepad++7.3.1
Easy-to-use and free code editor

SublimeText3 Chinese version
Chinese version, very easy to use

Zend Studio 13.0.1
Powerful PHP integrated development environment

Dreamweaver CS6
Visual web development tools

SublimeText3 Mac version
God-level code editing software (SublimeText3)

Hot Topics



Imagine an artificial intelligence model that not only has the ability to surpass traditional computing, but also achieves more efficient performance at a lower cost. This is not science fiction, DeepSeek-V2[1], the world’s most powerful open source MoE model is here. DeepSeek-V2 is a powerful mixture of experts (MoE) language model with the characteristics of economical training and efficient inference. It consists of 236B parameters, 21B of which are used to activate each marker. Compared with DeepSeek67B, DeepSeek-V2 has stronger performance, while saving 42.5% of training costs, reducing KV cache by 93.3%, and increasing the maximum generation throughput to 5.76 times. DeepSeek is a company exploring general artificial intelligence

AI is indeed changing mathematics. Recently, Tao Zhexuan, who has been paying close attention to this issue, forwarded the latest issue of "Bulletin of the American Mathematical Society" (Bulletin of the American Mathematical Society). Focusing on the topic "Will machines change mathematics?", many mathematicians expressed their opinions. The whole process was full of sparks, hardcore and exciting. The author has a strong lineup, including Fields Medal winner Akshay Venkatesh, Chinese mathematician Zheng Lejun, NYU computer scientist Ernest Davis and many other well-known scholars in the industry. The world of AI has changed dramatically. You know, many of these articles were submitted a year ago.

The performance of JAX, promoted by Google, has surpassed that of Pytorch and TensorFlow in recent benchmark tests, ranking first in 7 indicators. And the test was not done on the TPU with the best JAX performance. Although among developers, Pytorch is still more popular than Tensorflow. But in the future, perhaps more large models will be trained and run based on the JAX platform. Models Recently, the Keras team benchmarked three backends (TensorFlow, JAX, PyTorch) with the native PyTorch implementation and Keras2 with TensorFlow. First, they select a set of mainstream

Boston Dynamics Atlas officially enters the era of electric robots! Yesterday, the hydraulic Atlas just "tearfully" withdrew from the stage of history. Today, Boston Dynamics announced that the electric Atlas is on the job. It seems that in the field of commercial humanoid robots, Boston Dynamics is determined to compete with Tesla. After the new video was released, it had already been viewed by more than one million people in just ten hours. The old people leave and new roles appear. This is a historical necessity. There is no doubt that this year is the explosive year of humanoid robots. Netizens commented: The advancement of robots has made this year's opening ceremony look like a human, and the degree of freedom is far greater than that of humans. But is this really not a horror movie? At the beginning of the video, Atlas is lying calmly on the ground, seemingly on his back. What follows is jaw-dropping

Earlier this month, researchers from MIT and other institutions proposed a very promising alternative to MLP - KAN. KAN outperforms MLP in terms of accuracy and interpretability. And it can outperform MLP running with a larger number of parameters with a very small number of parameters. For example, the authors stated that they used KAN to reproduce DeepMind's results with a smaller network and a higher degree of automation. Specifically, DeepMind's MLP has about 300,000 parameters, while KAN only has about 200 parameters. KAN has a strong mathematical foundation like MLP. MLP is based on the universal approximation theorem, while KAN is based on the Kolmogorov-Arnold representation theorem. As shown in the figure below, KAN has

Regarding Llama3, new test results have been released - the large model evaluation community LMSYS released a large model ranking list. Llama3 ranked fifth, and tied for first place with GPT-4 in the English category. The picture is different from other benchmarks. This list is based on one-on-one battles between models, and the evaluators from all over the network make their own propositions and scores. In the end, Llama3 ranked fifth on the list, followed by three different versions of GPT-4 and Claude3 Super Cup Opus. In the English single list, Llama3 overtook Claude and tied with GPT-4. Regarding this result, Meta’s chief scientist LeCun was very happy and forwarded the tweet and

The latest video of Tesla's robot Optimus is released, and it can already work in the factory. At normal speed, it sorts batteries (Tesla's 4680 batteries) like this: The official also released what it looks like at 20x speed - on a small "workstation", picking and picking and picking: This time it is released One of the highlights of the video is that Optimus completes this work in the factory, completely autonomously, without human intervention throughout the process. And from the perspective of Optimus, it can also pick up and place the crooked battery, focusing on automatic error correction: Regarding Optimus's hand, NVIDIA scientist Jim Fan gave a high evaluation: Optimus's hand is the world's five-fingered robot. One of the most dexterous. Its hands are not only tactile

Target detection is a relatively mature problem in autonomous driving systems, among which pedestrian detection is one of the earliest algorithms to be deployed. Very comprehensive research has been carried out in most papers. However, distance perception using fisheye cameras for surround view is relatively less studied. Due to large radial distortion, standard bounding box representation is difficult to implement in fisheye cameras. To alleviate the above description, we explore extended bounding box, ellipse, and general polygon designs into polar/angular representations and define an instance segmentation mIOU metric to analyze these representations. The proposed model fisheyeDetNet with polygonal shape outperforms other models and simultaneously achieves 49.5% mAP on the Valeo fisheye camera dataset for autonomous driving
