The Future of AI
In the last decade, artificial intelligence (AI) has rapidly evolved from a theoretical concept to a practical tool that is already transforming many industries.
However, the true potential of AI is yet to be fully realized.
With trillions of dollars on the line, developments in the coming years will provide one of the most significant economic opportunities of our time. But as with any transformative technology, there are also challenges and risks ahead.
In this report, we’ll examine the future of AI in four critical sectors. Plus, we’ll take a closer look at potential regulations that could impact AI systems across the globe in the near future.
AI in Software Development
Many software developers have already seen huge productivity gains thanks to AI.
Not only can AI generate code, but it can also review and test code with the goal of detecting bugs and errors.
Currently, utilizing AI can make the software development process at least 25% faster.
Some research suggests the process can go as much as 55% faster with AI.
Using an AI coding platform such as GitHub’s Copilot can lead to faster completion of coding tasks and less attention to repetitive tasks.
OpenAI has recently proven its commitment to creating an all-encompassing AI solution for software development.
In the final months of 2022, the company hired approximately 400 computer programmers who are responsible for creating data to train AI models for software engineering tasks.
According to reports, the software developers are training the AI model on human-generated, step-by-step instructions for solving coding problems.
OpenAI’s current software coding model is called Codex.
Codex is the power behind GitHub’s Copilot, the most popular code-generating AI platform right now.
Search volume for “GitHub Copilot” increased rapidly in 2023.
Software engineers enter prompts or begin a chunk of code, and the platform automatically creates what they need to complete the code.
More than 40% of all the code written on the platform is AI-generated.
Company officials predict that will increase to 80% in the next five years.
In March 2023, the platform announced the release of Copilot X, an upgraded version that utilizes OpenAI’s GPT-4 and offers a wide range of additional AI capabilities.
The platform is GitHub’s “vision for the future of AI-powered software development.”
Copilot X offers several additional AI-powered capabilities for developers.
But just a month later, Amazon shook up the market when it announced its CodeWhisperer AI tool would be available free of charge.
Smaller companies are impacting the market too.
Search volume for “Tabnine” has exploded in recent years.
Company estimates show more than 4 million lines of code are written every day on the Tabnine platform.
Magic AI is a startup company looking to challenge CoPilot’s dominance in the future.
Magic AI says their platform will be like having a colleague inside the computer to assist with coding tasks.
The company closed a $23 million Series A funding round in early 2023 to bring their total funding to $28 million.
Despite the buzz, their actual platform isn’t available yet.
AI in Finance
In a recent Gartner survey, 64% of CFOs said they believe autonomous finance will be realized within the next six years.
This includes everything from customer service chatbots to automated forecasting to AI-powered fraud detection.
With AI, tasks like fraud monitoring, customer support, and personalized marketing can be automated.
Many companies are already moving toward that goal.
A 2023 survey showed that 75% of CFOs are planning to increase their technology spending this year.
The vast majority of CFOs are increasing their reliance on technology solutions.
The future cost-saving benefits of AI could be massive for the finance industry.
Just considering the middle-office tasks of banks, Insider Intelligence predicts AI will save these institutions $70 billion by 2025.
Data from NVIDIA shows most AI interest in the financial sector is focused on natural language processing, AI-powered recommendations for customers, portfolio optimization, and fraud detection.
More than 25% of finance executives are interested in using natural language processing platforms.
For financial institutions that want to make sense of large amounts of data, NLP is the ideal solution.
NLP platforms can analyze data and create insights or assess risk.
It can also be used to detect fraud, because AI can investigate certain words or phrases that bots are likely to use.
In the future, AI will likely be used not just to decrease costs but to increase profits for financial institutions as well.
For example, Zest AI offers an AI platform focused on lending.
Their AI helps banks spot good borrowers and avoid those who are likely to default.
Company data shows Zest clients see up to 30% more approvals with no additional risk when using their system to approve borrowers.
One credit union brought in $11 million in extra profit per year after implementing Zest’s platform.
AI-powered lending benefits banks and borrowers.
Portrait Analytics is another company looking to drive bottom-line results in the finance industry.
The startup’s offering is an AI investment analyst.
It’s a NLP platform that’s based on generative AI.
Analysts usually spend hours reading and analyzing company filings, but this AI application extracts that data automatically and summarizes the key information.
Portrait Analytics’ AI platform functions much like ChatGPT but focuses on investment data.
This allows firms to save time and realize new investment opportunities before the competition.
The company recently closed on $3 million in pre-seed funding.
AI in Healthcare
Goldman Sachs reports that more than 25% of the work done by healthcare professionals in the United States could currently be automated by AI.
Goldman Sachs estimates 28% of work done by healthcare practitioners could be automated by AI.
Researchers from McKinsey and Harvard estimate the AI-enabled cost savings could top $360 billion per year within the next five years.
There’s already a history of AI being used for things like radiology, pathology, and patient monitoring.
It was even used as part of the development of COVID vaccines.
In total, more than 500 AI algorithms have already been approved by the FDA. These are AI applications that directly impact patient care.
The vast majority of current AI applications are in radiology.
The first algorithm was approved in 1995, but there wasn’t much development until 2019. Between that year and 2022, there were 300 algorithms approved.
And, new AI developments are surging.
Precision medicine is one potential application of AI.
This refers to each patient being treated based on individual, unique parts of their DNA, medical history, imaging, and so on.
AI platforms will be able to analyze all of the patient’s data and recommend a tailored course of treatment specifically for that patient.
Today, this exists in the form of treating patients based on specific gene expressions, like certain cancers.
But experts say the AI technology will soon encompass more data points and be applicable to diseases like Alzheimer’s, obesity, and depressive disorders.
Precision medicine is one promising way to use AI to treat diseases, medical experts suggest.
In one current example, Avenda Health has launched an AI platform for managing and treating prostate cancer.
The solution combines patient data, imaging, biopsies, and pathology into a deep-learning algorithm. It then determines the extent of the cancer and creates a 3D cancer estimation map along with optimal treatment possibilities.
In clinical trials, use of the platform resulted in treatment changes 28% of the time. Most often, these changes resulted in more localized treatment of the cancer.
The Unfold AI system improves location mapping and treatment of prostate cancer.
Some in healthcare believe that AI may even go as far as replacing physicians in many aspects of the job.
Researchers from The Kellogg School of Management at Northwestern University say this transformation may be as little as a few decades away. But, they caution, there will still be a need for empathy, compassion, and human interaction from doctors.
Another future application of AI in healthcare is in pharmaceutical development.
Search volume for “AI drug discovery” is up 300% in the past five years.
It’s a $50 billion opportunity, according to Morgan Stanley.
The technology has the capability to dramatically increase the speed of getting drugs into the hands of patients.
One of the main ways in which AI does this is by mining the massive amounts of data created in the development process.
Consider that AI could sort through all of this data in a fraction of the time that it takes human researchers to do the same.
AI could notice a specific molecule that unlocks a new treatment or link data from various studies to come up with better drugs.
It can also be used to conduct human-like trials in the place of animal trials.
Using AI in pharmaceutical development is increasing the efficiency and speed of the entire process.
That’s the case for a potential ALS drug created by Verge Genomics through AI.
The company’s AI is able to utilize human data to create human models on the platform and use those to test the drug.
Verge’s drug, along with a few other AI drugs, is already hitting human trials.
Search volume for “Exscientia” has jumped 800% in the past five years.
In 2022, the company spent $155.6 million on AI drug development. That’s compared to just over $53 million in 2021.
AI and the Environment
In the future, questions surrounding AI and its impact on the environment will likely continue to swirl.
On one hand, AI applications could improve sustainability and lead to innovative environmentally-friendly practices.
However, the technology behind AI requires massive amounts of computational resources, which are characteristically damaging to the environment.
AI computing demands are growing quickly.
Google reportedly uses 15% of its total energy consumption to run machine learning workloads.
Other estimates say it takes an entire load of coal from a railcar just to train a LLM.
Less than 25% of AI systems run on clean energy.
Unfortunately, the negative impact on the environment begins long before the AI system is up and running. The manufacturing of the AI solutions is also problematic.
Manufacturing computer chips requires huge amounts of energy, uses billions of gallons of fresh water, and discharges hundreds of tons of hazardous waste.
But AI has helped society make incredible strides toward a more sustainable future, too.
PwC estimates that using AI for sustainability applications could reduce global greenhouse gas emissions by 4% by 2030.
AI applications have the potential to increase GDP by 4.4% by 2030.
The technology is utilized for making predictions about the environment based on a huge amount of data. AI can see trends and patterns that are nearly impossible for humans to see.
On a more industry-specific level, AI technologies can have a day-to-day impact on sustainability.
AI can enable agriculture sensors so farmers use less water, help restaurants identify and prevent food waste, and decrease energy use in factories, just to name a few examples.
The technology is also instrumental in the clean energy sector.
Search interest in “energy transition” is up nearly 500% in the past five years.
In the wind energy sector, AI can identify misalignment in the turbines and provide data that enables operators to fix the issues without requiring a worker to climb the turbine. It can also identify when parts are wearing out, up to six months in advance.
AI is also influencing the EV battery market.
As EV adoption grows, the need for copper, nickel, cobalt, and lithium is soaring—a $12 trillion supply gap is expected by 2050.
By 2050, experts predict there will be a huge supply gap of metals needed for EV batteries.
At the same time, the mining industry is making fewer mineral discoveries these days and are spending three times as much on finding new deposits compared to 30 years ago.
AI-powered mineral discovery aims to change this process.
KoBold’s AI system analyzes vast data sets like maps, measurements, and satellite imagery. Then their algorithms predict where they’re most likely to find minerals and guide their decisions throughout the mine discovery process.
Future Regulation of AI
The regulation of AI is a factor that many will be watching in the coming months and years.
Search volume for “AI regulation” has climbed 1,900% in recent years.
An Accenture study found that only 35% of consumers trust how companies are deploying AI, and 77% believe organizations should be held accountable for misusing the technology.
One tech policy group, the Center for AI and Digital Policy, urged the FTC to block the release of ChatGPT-4.
They claim the AI presents risks to businesses, individuals, and public safety.
The Center for AI and Digital Policy is advocating for increased regulation and oversight of AI.
Elon Musk and other tech executives have also brought up similar concerns, asking for a six-month pause on the release of ChatGPT-4.
ChatGPT-4 was released in early 2023 as scheduled, but that hasn’t calmed the waters.
Globally, EU officials have called for a summit so that world leaders can discuss the future implications of generative AI and potential regulation.
These are just a few examples pointing to potential regulation of AI and the push for responsible AI.
Search volume for “responsible AI” is up 1,400% since 2019.
The key tenets of responsible AI center on developing and deploying the technology in a fair and trustworthy way.
It also means that organizations ensure transparency, minimize bias, protect data, and provide a way for individuals to raise concerns.
Organizations deploying responsible AI are watching for potential bias and promoting transparency.
This type of accountability is one area of focus for potential AI regulations.
The U.S. government has said it’s particularly interested in overseeing AI’s impact on national security and education.
In April 2023, the Commerce Department asked for public comment on potential AI regulations, and Congress is reportedly working on legislation to regulate the accountability, transparency, and security of AI systems.
Debate surrounding the regulation of AI is likely to continue for several months or years.
Still, some in the industry say government regulation isn’t needed. They believe self-regulation of the AI industry is the way forward.
And, they warn that government regulations could slow progress and set the United States up to fall behind China in terms of dominating the future global technology.
As AI continues to advance, it will not only change the way companies operate but also the way individuals live. From personalized healthcare to automated investing, AI has the potential to revolutionize nearly every aspect of business.
Of course there are many challenges associated with the future AI technology. But there are countless opportunities available for bold innovators.
The future of AI is certain to be a rapidly evolving landscape.