With artificial intelligence (AI) standing front and center of emerging technologies, the state of AI in 2022 has become a hot topic. These discussions are fueled further by the popularity of generative AI, which can create text and image content with just a few words and pose a challenge to human content creators.
This calls for those interested in modern technologies to learn more about AI and its overall status in terms of advancements, breakthroughs, and availability. This lets you understand exactly where AI stands with its progress and where it can go from its current state.
To help you familiarize yourself with the latest AI trends, here is a lowdown on the current state of AI across different fields.
Compiled by Nathan Benaich and Ian Hogarth, the State of AI report summarizes the current status, progress, and standing of AI technologies across various fields. The report comprises trends in AI development as well as AI adoption, while also shedding light on what is driving the process of research within this world-changing segment.
Besides helping you learn more about how far AI has come in different segments, the report also explains the influence of key players in the AI sector. In addition to highlighting the most sought after aspects of artificial intelligence, it tells you about the considerations for safety in the AI space.
Some of the key takeaways in the state of AI report for 2022 include:
- The advancement of AI through smaller labs and engineering teams.
- The push for AI safety across different organizations.
- The adoption of AI across developed nations.
- The risks of AI being able to create harmful applications.
- The value of AI research across the world.
By going through the following findings on the current state of AI, you can get a firm grasp on where AI stands with its capabilities, what type of risks it brings to the table, and how it can drastically change the face of the world as we know it. This may allow you to fully appreciate the power of AI and how you can harness it for your own goals.
Since AI research requires extensive compute and data, it was speculated that only those labs with significant funding and resources would develop breakthrough research for the state of AI in 2022. This idea was further supported by the rising costs of computer hardware in 2021 that affected the retail as well as corporate sector.
But in the past year, this idea has been proven wrong by emerging labs with fewer resources than their well-funded counterparts. With entities such as Midjourney and Stability.ai becoming major players in the generative AI space, it has become clear that teams working on a budget can still achieve incredible results in this segment.
According to the State of AI report for 2022, Midjourney, which started from comparatively humble beginnings as opposed to its more resourceful counterparts, is already turning a profit by selling its model through a subscription plan. On the other hand, Stability.ai took under $600,00 to train its model and has become one of the most popular players in the segment as opposed to labs that take millions to achieve the same result.
With generative AI rising to new heights of popularity through labs such as OpenAI and stabilityai, the risks of AI being used for unsavory purposes have become higher than ever. Since anyone can generate harmful content through the directive of a few words, it is crucial that those who offer AI tools pay attention to preventing these attempts.
This has also pushed the current state of AI to include safety practice at the core of AI models, algorithms, and programs. Besides ensuring that no one can generate content that is offensive or damaging, this also refers to training AI through models that are free of dangerous content. This includes but is not limited to content that revolves around racism and misinformation.
This has led major AI labs to expand their safety research teams by a whopping 200 percent as compared to 2021. In 2022, nearly 300 safety researchers were employed by these organizations, which is a major increase from the 100 safety researchers that major labs employed a year ago. With the number of safety experts increased, the sector seems to be preparing itself for handling the many risks that AI brings to the table.
While harboring friendly relations when it comes to trade, the U.S. and China have been at odds across many other aspects. This also includes the challenge to advance in emerging technologies before the other. This holds true for the state of AI in 2022 as well, which sees both nations in a race to surpass the other in the usage of this technology.
In 2022, U.S. based researchers published more than 10,000 papers in the subject of AI. Whereas, Chinese researchers published approximately 7,000 papers. While American researchers increased their output by 11 percent as compared to 2021, Chinese researchers did so by 24 percent for the same period. But these statistics are limited to papers published in the English language.
When you broaden the perspective and look at the research papers published in the Chinese language, Chinese researchers surpass the output by U.S. researchers who publish their papers in English. Since 2010, Chinese researchers have published approximately 254,000 papers as compared to the approximate 57,000 papers published by American researchers.
There has been a lot of talk about the current state of AI becoming proficient enough to generate creative content and taking jobs away from artists, writers, and other creators. The ethics of producing AI art, texts, and other works have sparked a heated debate across various sections of the internet. But the risk of AI being advanced enough to produce biological weapons somehow goes unnoticed.
One of the most beneficial yet often ignored use cases of modern AI is the ability to identify patterns across complex datasets for drug discovery, which has led 18 of AI-first drug discovery patterns to be up for clinical trials by 2023. This could lead to different types of molecular data being processed to discover cures for various conditions. But when used under reverse engineering, this has great risks for the same data being used for the creation of biological weapons.
Similar to many other segments of AI, the applications for drug discovery are experimental and their results are not always completely accurate. But with time and technological advancements, these models may become adept enough to not only find new cures but also develop the path to new diseases in a rather seamless manner.
When discussing the state of AI in 2022, it is difficult to overlook the role that the U.S. plays in the active advancement of this technology. In 2022, AI investment in the U.S. stood as more than half of the world’s venture capital in the sector. This highlights that the U.S. has become the hub for AI research and continues to lead the charge on practical applications.
In 2022, the U.S. led through the creation of 292 unicorns, which are companies with a total valuation of over $1 billion. Through these unicorns, the U.S. has created approximately $4.6 trillion of value. On the other hand, China trails behind with 69 AI unicorns, while the U.K. holds 24 unicorns to its name.
With the current state of AI, the U.S. remains one of the undoubted winners of the sector in 2022. With projects such as OpenAI, which has created amazing AI models including Dall-E 2 and GPT-3, it is not difficult to imagine that the country may continue to support AI-based ventures to create more value for the sector.
Even if you take a fleeting look at the state of AI in 2022, it is not difficult to determine that the technology is advancing at an exponential rate. Apart from its ability to perform niche analyses and activities, AI has also proven quite adept at carrying out certain actions for individuals and their required activities.
From content generation to personal assistants, AI has opened doors to a wide variety of applications in various sectors. Seeing these capabilities, you can easily project that AI may redefine the way that we do things on a day to day basis. Whether or not you want to give the technology a chance yourself is another matter altogether.
The full State of AI report from 2022 is available here. By going through the detailed findings, you can learn more about what to expect from this incredible subset of technology.