How artificial intelligence and machine learning will reshape enterprise technology | Tech Rasta


Artificial intelligence (AI) and machine learning (ML) are ubiquitous in consumers’ lives, from the “next” suggestions from your streaming service to the routes your GPS suggests when you plug an address into your phone for directions. Less visible impacts of AI and ML include using AI to control data center capacity and cooling, or managing restaurant wait times, with some companies using AI to make decisions about how many burgers to cook for the day’s lunch rush.

While AI refers to a computer’s ability to mimic human decision-making, ML is the algorithm-based foundation that enables AI. We can think of automation as the application of AI to develop a series of repetitive tasks or actions designed to complete a specific task or execute a process. Companies use automation to ship products to warehouse workers for packing, processing invoices, and assisting with many other repetitive business tasks that have historically been performed by humans. Likewise, many people who have worked in an office have probably seen automation show up in a PDF program in the form of a “detect text” feature.

Although automation is not yet as common as AI and ML, current economic conditions will increase its prevalence. With inflation at 9.1 percent in June, rising labor costs and the Federal Reserve raising interest rates further, businesses across all sectors need to explore AI, ML and automation applications as ways to improve speed and efficiency while reducing costs.

What is the difference between artificial intelligence and machine learning?

While AI refers to a computer’s ability to mimic human decision-making, ML is the algorithm-based foundation that enables AI. We can think of automation as the application of AI to develop a series of repetitive tasks or actions designed to complete a specific task or execute a process.

More from Nate FarschiWho’s Going to Pay for American-Made Semiconductors?

What are the enterprise applications of automation?

Even in ideal economic conditions, companies still want to explore AI, ML and automation in their relentless pursuit of greater efficiency. But these investments will become even more critical as the economy slows down.

According to Gartner, the different types of automation fall into three basic groups: task automation, process automation, and augmentation. Some important sub-categories include Robotic Process Automation (RPA), Business Process Automation (BPA), Conversational AI, Natural Language Processing (NLP) and Optical Character Recognition (OCR).

AI, ML and automation offer technology-based solutions to time-consuming business challenges traditionally handled by entry-level workers. These solutions can significantly reduce implementation time, improve accuracy, reduce costs in the long run, and free up employees to add value to their organizations in other ways.

According to an International Data Corp. survey published by IBM, the leading enterprise use cases for AI in 2020 are IT, human resources, customer service agents, business processes, and automation in the areas of IT threat intelligence and prevention. Respondents to a special report from BPTrends that same year indicated that “nearly 75 percent of respondents believe (business process management) and technology have helped their organizations achieve their goals.”

A bar chart showing the effectiveness of automation
Image created by RSM US LLP

In our view, the potential for these technologies to help businesses achieve their goals will only increase from 2020. Beyond the five prominent enterprise use cases highlighted above, some companies are already looking to mitigate the effects of the current economic climate by automating expense reports. Employee onboarding, purchase orders, work orders, vacation requests and more, as explained in this article from workflow automation software company Frevvo.

More about Artificial IntelligenceWhat is Artificial General Intelligence?

How will automation affect investments?

Many strong signs indicate that AI, ML and automation will play an important role in the future of business. Currently, “only 12 percent of enterprise applications are using AI and ML at a level of maturity that results in a significant competitive advantage,” says Accenture. At the same time, IDC predicts that “90 percent of enterprise applications will be replaced by artificial intelligence and machine learning (AI & ML) by 2025.”

Private equity and venture capital deal activity associated with AI, ML and automation tells a slightly different story from the declines and withdrawals of the broader economy. According to Pitchbook data, in 2021, investors will put $249 billion into AI and ML companies and $101 billion into workflow automation companies globally. By the first quarter of 2022, pre-money valuations increased for all stages of VC investment, and median deal value increased for angel, seed and late-stage companies. IDC estimates the 2023 market size for systems using AI and ML to be just over $1 trillion at $979 billion. Gartner predicts the global market for automation technology to reach $596.5 billion by the end of 2022.

Combined venture capital (VC) and private equity (PE) firms hold nearly $1 trillion in dry powder capital. US private equity held $731.4 billion in dry powder as of March 2022, with middle market PE firms holding $423.1 billion, according to Pitchbook. As of June 2022, US venture capital firms held a record $290 billion in dry powder, a 23 percent increase from VC dry powder levels at the end of 2021. All told private equity and venture firms investing in AI & ML companies. Plenty of capital should be allocated.

Adoption of AI and ML applications is also becoming cheaper and faster. According to Fortune, technologies that cost more than $1,000 in 2017 will cost $7.43 in 2021, and processes that took 372 seconds in 2018 will take 47 seconds in 2020. According to Stanford, “Since 2018, the cost of training an image classification system has decreased by 63.6 percent, while training time has improved by 94.4 percent. The same report found lower training costs and faster training times in other parts of the Stanford AI competition.

Especially from an enterprise perspective, AI, ML and automation are making their application faster, more user-friendly and less expensive, while labor and debt are becoming more expensive. All the while, median deal sizes and valuations for AI, ML and automation companies haven’t taken a big hit as the big tech sector and US VC and PE firms sit in record dry powder. The future of AI, ML and automation looks bright and has a very high ceiling.


Source link