Artificial intelligence has become more prominent than ever before, with its development continuously increasing. Why is it receiving more attention? Due to the latest advancements in generative artificial intelligence, individuals, research teams, and companies are all striving to become leaders in this field, making the competition more intense. However, it can be difficult to keep up with all the latest trends.
To further summarize the trends in artificial intelligence for 2023, the AI Index Report has been released. Compiled by the Stanford Institute for Human-Centered Artificial Intelligence (HAI) in collaboration with a guidance committee from various academic and industrial domains, the report aims to provide decision-makers and audiences with responsible and ethical insights to advance artificial intelligence.
What are the trends and key findings of this report? Let's discuss.
"2023 AI Index Report"
The "2023 AI Index Report" is the sixth edition of the report, focusing on the development of artificial intelligence and its impact on human society. Based on data collected and rigorously analyzed from various experts, here are the key points highlighted in the report.
Industry Leading over Academia
According to the report, many significant machine learning systems released in 2022 were led by industry rather than academia. One notable machine learning system is the model included in the Epoch dataset, which has specific standards such as improving state-of-the-art methods.
In 2022, 38 important machine learning systems were released, with 32 of them being produced by the industry. The chart below illustrates the trend in 2022 compared to 2014.
Saturation of Traditional Benchmark Performance
Many machine learning models are tested annually on standard benchmark datasets to assess whether they can improve the current benchmark performance. For example, image classification tasks using the ImageNet dataset or English language understanding tasks using the SuperGlue benchmark.
However, the annual performance growth has become increasingly marginal, reaching a plateau. The chart below shows that many benchmarks improved by less than 5% year-on-year.
To alleviate benchmark saturation, researchers have recently introduced new comprehensive benchmarks such as BIG-bench and HELM. Although these new benchmarks mainly focus on large-scale language model tasks, as it is the current trend.
AI's Impact on the Environment
Recent studies have shown that the selection of large-scale language models can have a negative environmental impact by emitting carbon dioxide exceeding any other human activity. The chart below demonstrates that the carbon emissions of the GPT-3 model are nearly 500 times that of a regular passenger's air travel.
However, research has also shown that artificial intelligence can help save the environment. In 2022, DeepMind's research demonstrated that a machine learning agent called BCOOLER can effectively improve energy-saving systems within Google data centers by 12.7%.
Accelerating Scientific Progress with AI
In 2022, artificial intelligence has proven to contribute to many scientific breakthroughs, including:
1. DeepMind researchers created a reinforcement learning algorithm to find the optimal approach to managing plasma generated by tokamak systems.
2. Fawzi et al. (2022) discovered a new algorithm called AlphaTensor that uses reinforcement learning for matrix operations.
3. NVIDIA discovered a new method to improve chip systems, allowing artificial intelligence to design systems using reinforcement learning.
4. Shanehsazadeh et al. (2023) developed a new antibody discovery system using zero-shot learning with artificial intelligence.
Increasing Number of AI Misuse Cases
With the new developments in artificial intelligence, the number of misuse cases is expected to increase. According to the AIAAIC database, the number of incidents in 2021 was 2012 times higher than in 2012. The trend is shown in the chart below.
An example of controversy is the student emotion monitoring system developed by Intel.
The increasing number of cases also indicates a higher awareness of AI misuse, leading to improved tracking of such behaviors.
Growing Demand for AI-Related Skills
Over the years, there has been an increasing demand for AI-related skills in job postings across various countries. Data shows that many companies have shown greater interest in implementing artificial intelligence within their organizations. The Lightcast data below demonstrates the rising trend in job postings requiring AI-related skills.
The data also shows that the United States has the highest proportion among countries in terms of demand for AI skills, while New Zealand has the lowest. Although the data is limited to selected countries, caution is still necessary.
Decrease in Private AI Investments in 2022
According to NetBase Quid data, the total investment in AI-related ventures in 2022 amounted to $91.9 billion. Although this figure is lower than in 2021, it is still 18 times higher than in 2013. The chart below illustrates the data.
Despite the short-term decline, the long-term trend is still growing. This indicates that many companies are still willing to invest in the development of artificial intelligence.
Although the adoption rate of AI in companies tends to stabilize, those already adopting AI continue to make progress.
According to a McKinsey report, as of 2022, approximately 50% of surveyed companies have implemented artificial intelligence in their business functions. The chart below shows that the adoption rate has stabilized between 50-60% since 2019, although this number is significantly higher than in 2017.
This finding suggests that companies are still willing to adopt artificial intelligence in their business and organizations, although this proportion may remain around 50% in the coming years.
Increasing Interest in AI Legislation Institutions
According to the AI Index study of 127 AI legislation institutions in various countries, 31 countries have passed AI-related bills, with a total of 37 AI-related bills in these 123 countries. The number of AI-related bills in 2022 was 37, compared to only one bill in 2016. The trend is shown in the chart below.
The data above indicates that many countries have recognized the potential of artificial intelligence in various domains but require regulatory mechanisms to prevent misuse.
Global Recognition of AI Benefits
Through an IPSOS survey conducted in 28 countries with 19,504 adult respondents of different age groups, over 50% of people understand how artificial intelligence works and believe that it will change their daily lives in the next 3-5 years. The complete data can be seen in the chart below.
The high number of people who believe in the usefulness of artificial intelligence indicates that it will gradually integrate into people's daily lives. However, it is normal for some individuals to express anxiety about artificial intelligence.
Conclusion
The trends in artificial intelligence are on the rise and are likely to continue in the coming years. The AI Index Report confirms this trend, highlighting the following key points:
- Industry leading over academia
- Saturation of traditional benchmark performance
- AI's impact on the environment
- Accelerating scientific progress with AI
- Increasing number of AI misuse cases
- Growing demand for AI-related skills
- Decrease in private AI investments in 2022
- Continuous progress of companies adopting AI
- Increasing interest in AI legislation institutions
- Global recognition of AI benefits
Source: https://www.kdnuggets.com/2023/06/2023-ai-index-report-ai-trends-expect-future.html
Alphabet Q1 paid subscribers reached 350M, fueled by YouTube & Google One. YouTube ad revenue slightly missed at $9.88B amid Premium growth. Total Q1 revenue hit $109.9B, beating forecasts.
Anthropic releases Claude connectors for Adobe, Blender, and Ableton, enabling AI workflow automation and app control. It also sponsors the Blender Development Foundation.
Apple introduces an App Store model splitting 12-month plans into monthly installments. On iOS 26.4+, it lowers upfront costs without reducing total price, adding billing tracking.
Google launches Aadhaar digital IDs in Wallet for India, enabling secure QR/NFC verification. A phased partner rollout targets key sectors, while passport-based credentials expand globally.
YouTube tests "Ask YouTube," an AI tool blending text and video for complex queries. Initially for US Premium users, it supports interactive follow-ups, expanding Google's AI search.
Google grants US DoD classified AI access, joining OpenAI and xAI in defense deals amid Anthropic's dispute. Contractual safeguards against surveillance and lethal weapons remain unenforceable.
Mistral AI launches Workflows, a Temporal-powered orchestration platform for secure, Python-driven enterprise AI automation featuring human-in-the-loop approvals.