After the Billion-dollar Cloud Computing Project, the Pentagon Is Going to Spend Money Again

In recent years, the Pentagon has been making continuous moves in artificial intelligence, with multiple approaches such as investment, technology, and talent introduction. Recently, its artificial intelligence center JAIC is seeking testing tools and technologies for artificial intelligence products. This gives us a glimpse into the US Department of Defense's ambitions in AI.

Keywords: Pentagon artificial intelligence bidding

Recently, the Pentagon is going to bid again!

The previously highly anticipated multi-billion dollar cloud computing project was signed by Microsoft not long ago, and they quickly launched a new artificial intelligence project.

The billion-dollar order has just been signed, but the sponsor is still not satisfied

In March 2018, the Pentagon announced the Joint Enterprise Defense Infrastructure (JEDI) project to build a cloud computing platform to support military weapon systems and confidential data storage.

The project is worth as much as 10 billion US dollars, and this piece of "fat meat" has made technology giants such as Amazon, Microsoft, and Oracle eager to move.

At the beginning, Google was also involved, but because the Maven project caused joint opposition from employees, the relationship between Google and the Pentagon has always been delicate, which is equivalent to walking away from the table early.

The bidding results of the JEDI project have also been a focus of attention in the industry.

Last October, the Pentagon announced that Microsoft had won the contract, but Amazon strongly disagreed and immediately filed a lawsuit. However, the order was ultimately awarded to Microsoft.

Just as the 10 billion dollar order was finalized, the Pentagon released a new project. It seems that 10 billion dollars is not enough to solve the Pentagon's AI plan.

The Pentagon wants to add AI to the battlefield. People are stupid and have a lot of money. Come quickly

In June 2018, the Pentagon established the Joint Artificial Intelligence Center (JAIC) to accelerate the application of artificial intelligence in intelligence and military fields.

Recently, JAIC published a Request For Information (RFI) on the government bidding platform, stating that they are seeking cutting-edge artificial intelligence evaluation technology to ensure the safety and effectiveness of future AI products. The evaluation scope includes emerging AI technologies such as machine learning, deep learning, and neural networks.

In recent years, the Pentagon's demand for AI technology has continued to grow.

They believe that artificial intelligence has great potential to provide timely support for combat, improve the tempo of combat, and thus increase advantages on the battlefield.

Therefore, the website also clearly states that the purpose of this procurement is to enhance the ability of combat personnel to perform repetitive and dangerous tasks.

Pentagon: I also want an AI service that evaluates AI

The invitation letter stated that in order to ensure the production of safe and effective products, JAIC has a Test and Evaluation (T&E) office.

The JAIC T&E Office is responsible for evaluating all JAIC products, conducting algorithmic, system, and operational testing, and providing independent AI performance analysis.

The needs of the JAIC T&E team mainly include the following two aspects:

1. Testing techniques and tools

The JAIC T&E team is seeking testing technologies and tools for various AI applications, including machine learning, deep learning, and neural networks, with a focus on the following six areas:

1. Use voice to text conversational interface applications;
2. Products and services that provide voice support for Department of Defense applications and systems;
3. Image analysis, testing visual search and image classifiers based on deep learning;
4. Products and services that support natural language processing (NLP);
5. Humans augmented by machines, including human-machine interfaces and improved methods to measure cognitive and physical workloads on warfighters, including augmented reality (AR) and virtual reality (VR) testing services;
6. Autonomous Systems

2. Testing and Evaluation Services

The JAIC T&E team is also seeking services from government, academic, and industry organizations to assist with technology testing and evaluation provided by JAIC or other DoD staff.

Specifically, support is needed in the following five areas: dataset development/management, test tool development, model output analysis, test planning, documentation and reporting, and testing services.

AI has been developed for several years and cost hundreds of billions of dollars

In the past two years, the Pentagon's various initiatives have demonstrated the importance the U.S. Department of Defense attaches to artificial intelligence, including the establishment of multiple artificial intelligence management and guidance departments, ensuring priority investment in artificial intelligence, and the introduction of a national artificial intelligence strategy.

We briefly review some of the major initiatives of the U.S. Department of Defense in the field of artificial intelligence since the establishment of JAIC:

The above is only part of the Pentagon’s artificial intelligence layout, all of which reveal the U.S. Department of Defense’s urgent need for artificial intelligence technology.

Adding artificial intelligence to the battlefield has never stopped the controversy

In fact, the Pentagon’s emphasis on artificial intelligence is not something that has happened in the past two years. DARPA has been supporting research on artificial intelligence for more than 50 years, but it has only begun to focus on promoting progress in this area in recent years.

But the military use of artificial intelligence has been controversial from the beginning.

For example, in terms of ethics, should artificial intelligence be used as a lethal weapon? The Pentagon has also recruited "ethicists" to supervise the US military's AI and established responsible, fair, traceable, reliable and manageable AI ethical principles.

For example, on the battlefield, if an AI system gives an unprecedented command, should soldiers obey and execute it? Because fundamentally, people still lack confidence in machine reasoning, especially in the special environment of the battlefield, where variables and emergencies may occur at any time, and these situations are things that machine designers have never experienced before.

-- over--