Biden, sudden warning!

US President Biden is going to shoot artificial intelligence again!

On the evening of March 7, local time, US President Biden delivered a high-profile State of the Union speech in the election year, which was also his third State of the Union address during his tenure. Biden called on Congress to take urgent action to regulate artificial intelligence through legislation. Biden warned that this technology will bring risks to Americans if it is not controlled.

It is noteworthy that it has been reported recently that the EU is ready to formulate comprehensive artificial intelligence rules or have a global impact.

Look at the detailed report!

Biden made an appeal.

According to CNN quoted by the Global Times, US President Biden delivered a State of the Union speech in the House of Representatives on the 7th local time. In his speech on domestic issues in the United States, Biden called on Congress to take urgent action to regulate artificial intelligence through legislation, including prohibiting "artificial intelligence voice imitation."

Just a few weeks ago, there was an incident of "AI imitating Biden’s phone call" in the United States. The "fake Biden" phone call generated by AI called on 5,000 voters who are most likely to support the Democratic Party not to vote in the New Hampshire primary. CNN subsequently reported that the US government traced and found that 71-year-old man Walter Monk and his Texas-based "Life" company were behind the scenes.

On the 7th, Biden urged legislators to "use the promise of artificial intelligence to protect us from its dangers" and warned that this technology will bring risks to Americans if it is not controlled. However, CNN said that for about a year, American legislators have been working hard to promote any meaningful artificial intelligence legislation. Although false information experts warn of the threat of artificial intelligence to opinion polls and public discourse, few people expect Congress to pass legislation to control the artificial intelligence industry in a divided election year.

At the end of October last year, US President Biden signed an executive order on artificial intelligence, and introduced the first set of White House regulations on generative artificial intelligence. According to the executive order, many government agencies in the United States need to formulate standards to prevent the use of artificial intelligence to design threats such as biological or nuclear weapons, and seek the best method for content verification such as "watermarking" and draw up advanced network security plans. Specifically, the order requires testing artificial intelligence products and reporting the test results to the federal government. It also proposes to attract global artificial intelligence talents to stay in the United States.

The White House said in a statement at the time that the executive order aims to ensure that the United States plays a leading role in fulfilling the promise of artificial intelligence and managing risks. As part of Biden’s comprehensive strategy of responsible innovation, it is based on the actions taken by the President before, including the voluntary commitment of 15 leading companies to promote the safe, reliable and trustworthy development of artificial intelligence.

In addition to artificial intelligence, Biden also mentioned a number of topics in his State of the Union address speech on March 7. In his speech, Biden reiterated that he would not send troops to Ukraine, but at the same time stepped up his appeal for assistance to Ukraine and targeted his Republican rival, former President Trump.

Biden vowed not to send American soldiers to Ukraine, saying that Ukraine wants military aid and weapons to fight back against Russia, not American personnel. "They don’t want American soldiers. In fact, there are no American soldiers fighting in Ukraine. I decided to keep this state. " Biden said. He stepped up his appeal to the United States to continue to provide assistance to Ukraine. Biden said that if the United States ignores Ukraine now, "Ukraine will face danger and Europe will face danger."

In his speech, Biden criticized Trump, who had previously encouraged Russia to "do whatever they want" to NATO countries that failed to fulfill their defense spending obligations. Biden said, "I think this kind of speech is outrageous, dangerous and unacceptable." CNN said that Biden did not directly mention Trump’s name, but called him "my predecessor, the former president of the Republican Party".

On the current Palestinian-Israeli conflict, Biden said that after the attack on October 7 last year, Israel has the right to attack Hamas, but it also has the "basic responsibility" to protect Gaza civilians. He said that he would instruct the US military to lead an urgent task to build a temporary dock on the coast of Gaza to receive a large amount of humanitarian supplies.

Biden also made comments on many issues such as education and economy.

The EU has also made big moves.

It is worth noting that in the field of artificial intelligence, the EU will also make big moves: comprehensive artificial intelligence regulations will be adopted soon.

Recently, foreign media reported that the EU is about to adopt comprehensive artificial intelligence regulations, which are very strict. The European Parliament is scheduled to vote on this legislation on March 13th, and a member of parliament said that he expected the legislation to be passed easily. According to the current timetable, the listing bill will basically take effect in 2026, and some provisions will take effect this year.

The bill will regulate different uses of artificial intelligence according to risks, including banning certain use cases, such as the "emotion recognition" system in the workplace, which will face the highest fine: 7% of global income or 35 million euros, whichever is higher.

The report pointed out that the EU will not be the first jurisdiction to regulate artificial intelligence, but the size of the European market means that the EU Artificial Intelligence Act will be affected globally, affecting the way multinational companies manage data and use artificial intelligence. The law will apply to enterprises operating in the markets of 27 countries and builders of artificial intelligence systems used by the European Union headquartered anywhere, including the United States.

In the United States, legislators also discussed artificial intelligence supervision, held industry hearings, and introduced bills. However, Congress has not yet passed legislation to widely regulate the use of this technology.

Frequent explosion models in the industry

At present, for artificial intelligence with great potential, regulatory legislation is one aspect, and development is also a very important aspect.

Recently, Sora detonated a new round of AI arms race, and the era of general artificial intelligence accelerated. In mid-February, Google released Gemini 1.5 Pro, which enhanced its ability to handle a large number of videos, texts and images. Developers can submit up to 1 million tokens to Gemini 1.5 Pro, which is equivalent to about 1 hour of video, 11 hours of audio or 700,000 words of text data.

Immediately, OpenAI released the Wensheng video model Sora, which can generate high-definition smooth video up to 60 seconds with text instructions, and has significant advantages in generating video length, coherence and multi-shot switching. Compared with the previous Wensheng video models Runway, ika and Stable Video, it has achieved cross-generation upgrade.

Galaxy Securities believes that the iterative speed of overseas technology giants to upgrade multimodal large models far exceeds expectations, Sora can simulate actions that affect the state of the world in a simple way, and the era of general artificial intelligence may come ahead of schedule. The broker further pointed out that multi-modal scenarios such as Sora and Gemini will broaden the downstream application scenarios, which will provide new production tools for movies, advertisements, games, short videos and other fields, and related industries are expected to reduce costs and increase efficiency and take the lead in application.

Sora’s powerful Wensheng video capability will rapidly promote the commercialization of the application end, stimulate the industrial scale, and then transmit it to the upstream computing infrastructure, and the demand for computing infrastructure will once again erupt. According to the calculation, Sora’s demand for computing power at the reasoning end will increase exponentially, which will drive the demand for infrastructure such as upstream chips, servers, optical modules and liquid cooling.

In addition, on March 5th, Anthropic released a new generation of AI big model series -Claude 3. There are three models in this series, which rank from weak to strong: Claude 3 Haiku, Claude 3Sonnet and Claude 3 Opus. Claude3 Opus has surpassed GPT-4 and Gemini 1.0 Ultra in many evaluation benchmarks, showing strong computing power in many fields, and further promoting the benchmark of artificial intelligence big model industry. Guolian Securities said that the Claude3 model was released, and the development of artificial intelligence continued to be interpreted, focusing on the three main lines of computing power, model and application.

Dongxing Securities said that the current artificial intelligence industry support policies and explosion models are frequent. As one of the engines of new productivity, artificial intelligence may promote the upgrading of traditional industries, and it will continue to make breakthroughs in the application side under the background of digital industrialization and industrial digital development. At the same time, it will benefit from the dimensions of data and landing in education, medical care, law and other fields. At the same time, the development of multi-modal and the release of new models will bring about the improvement of computing power demand and the pattern change on the model side. Model companies and application companies are expected to benefit from the development of the industry.

In the secondary market, artificial intelligence concept stocks continue to be active. On March 7th, in the US stock market, NVIDIA, the global leading AI computing enterprise, rose by more than 4%, reaching a new high for six consecutive trading days, with a market value close to 2.3 trillion US dollars, less than 300 billion US dollars from Apple’s closing market value. In the A-share market, on March 8, Industrial Fulian, Gaohong, Shengtong, and CLP Xingfa had daily limit.

Editor: Wang Lulu

Proofreading: Zhu Tianting