AI will thrive in 3 key areas in 2023
Some of the biggest tech names have laid off artificial intelligence (AI) and machine learning (ML) employees this fall, including Meta, Twitter and Amazon. In light of that, it would make sense for industry nerves to be high entering 2023, but that’s not the case.
Even in the midst of a possible recession, AI experts across several industries told VentureBeat that they expect AI innovation to continue and companies to adjust budgets and priorities accordingly. In fact, these industry leaders resoundingly underscored three areas where AI has thrived over the past year and will continue to grow in 2023: workplace automation and human-centric AI; data-driven AI decision-making; and generative AI use cases.
“The U.S. is knee-deep in a recession, but despite the economic uncertainties that consumers and companies are facing, I predict AI will remain not only resilient but recession-proof,” said Scott Stephenson, CEO at Deepgram, an AI-powered transcription tool. “AI will continue to be central to business in 2023, by cutting costs and increasing innovation. Simply put, AI will help us do more with less.”
While some may hear “cutting costs and increasing innovation” as workplace automation and artificial intelligence tools pushing them out of a job, several leaders say that what’s more likely is automation taking over mundane tasks, so employees can focus on more advanced ones.
“AI won’t — and shouldn’t — replace humans in the near term,” said Vishal Sikka, founder and CEO of human-centered AI platform, Vian AI. “The reality is that AI today, as impressive and as powerful as it is, is nowhere close to human judgment. In 2023, the recognition that too many platforms aren’t designed for humans to use will increase. More and more systems will be designed to amplify human judgment — to aid people and encourage AI symbiosis, rather than seeking to have AI replace the user.”
Human-machine interactions need human-centric solutions
As artificial intelligence continues to shift how we work and the tools we use for work, several experts say the tools themselves should become more human-centric too — simplifying things in day-to-day workflows, or creating a platform that eases two-way human and technology interactions.
“We’ll see the proliferation of user-friendly, non-technical AI systems,” said Zeeshan Arif, founder and CEO at software development company Whizpool. “They’ll be building more AI systems internally to help them streamline their operations and improve their customer service.”
Pieter Buteneers, director of engineering in ML and AI at Sinch, a cloud communications platform, echoed this, telling pub VB that heading into the new year, AI will likely move away from keywords as it “progress[es] toward actual comprehension and understanding.”
Advancements in natural language processing (NLP) and large language models (LLMs) are likely to occur in this vein as well, industry leaders told VB, noting that these technologies can assist with scaling business processes.
“NLP is revolutionizing how humans interact with machines; these technologies can understand what people say, act on that information appropriately and respond accordingly,” said Devanshu Bansal, cofounder of The X Future, an ideation platform for corporate teams. “NLP has a lot more to offer than just clearly communicating with users; it can also help scale business operations by helping them understand the voice of the customer.”
On top of that, with increasing advancements toward human-centric AI, and particularly in NLP and LLMs, Julien Salinas, founder and CEO at NLP Cloud, an AI startup that deploys LLMs, predicts that this progress may even make the technology more affordable.
“We keep seeing a stable growth of adoption of AI technologies when it comes to natural language processing,” Salinas said. “I expect large language models to become cheaper thanks to many low-level optimizations being made by the AI community right now.”
AI and ML models will be driven by quality data
Although AI has experienced rapid growth throughout the past year, some say it is not advancing quickly enough, getting stuck in a “Stone Age” of sorts — which is why so many ML projects fail. Looking ahead to 2023, this is something industry leaders are increasingly aware of and aiming to address.
Sameer Maskey, CEO of Fusemachines, an AI educational platform, and professor of AI at Columbia University, told VB that “at the enterprise level, data silos continue to present a big obstacle…organizations are slowly beginning to understand that the success of AI hinges on data, and a lot of it.”
As a result, Maskey expects to see more solutions enabling access to “credible pools of data that will aid businesses to benefit from AI-powered efficiencies.”
One solution coming on the scene to address AI’s data-starvation is the use of synthetic data, which Gartner predicts will be used to accelerate 60% of AI projects by 2024. Another is the use of foundational models, which are typically trained on large amounts of unlabeled data and then paired with smaller sets of labeled data to propel problem-solving.
“The general trend towards data-centric AI seems to be accelerating,” said Ulrik Stig-Hansen, president and cofounder at computer vision company Encord. “This lowers the barriers to entry for companies to start monetizing their data, and all things [being] equal should lead to more widespread adoption of these technologies. Over the next year, companies will need to learn how to use data strategically. This is where they can really find the competitive advantage of AI.”
Generative AI’s use cases will explode
Advances in generative AI, which may be unofficially dubbed the “topic of the year” in the field, took off this year, giving rise to several new companies and tools that are revamping how creatives and non-creatives alike do their work. Tools like DALL-E, Imagen and Stable Diffusion generate original text-to-image concepts almost instantly after they are given even the most obscure of prompts — like “an AI bot sitting on a throne.”
On top of its growing use cases, Encord’s Stig-Hansen predicts that “there will be a transformational leap forward in the availability of generative AI.”
This domain shows no signs of slowing down in 2023, with Gartner predicting that generative AI will not only improve digital product quality, but by 2025 will also account for 10% of all data produced — compared to a current 1%.
Generative AI is a tool that, though not quite replacing jobs yet, is sparking curiosity across enterprise sectors about what might be ahead.
“As the technology grows more sophisticated it will continue to be disruptive, not just for images and content development, but for other industries like speech recognition and banking,” Deepgram’s Stephenson said. “I predict that generative technology will soon act as an exoskeleton for humans — it will support the work we are doing and ultimately drive a more efficient and creative future.”
With the text-to-image, text-to-video and upcoming text-to-3D capabilities that generative AI provides, adoption of the tool will also increase — probably even outside of the technology and enterprise spaces into others like entertainment and the creative professions, Stig-Hansen said.
“These AI models will only get better and become more photo-realistic,” he said. “The space will evolve from just AI-generated faces to whole bodies in both imagery and video. There will also be a huge adoption of generative AI in creative industries, including the music industry.”
Investors are also beginning to take note of the new technology, but some are waiting to see if it will remain all “hype” or if its growth becomes solid for a forward trajectory.
Bansal, the cofounder of The X Future, said their fund, Z Nation Lab, was an early investor in AI-powered content generation platform Jasper AI, and said he sees it only going up from here.
“Seeing the exponential growth that Jasper AI has achieved in the past two years makes me very bullish on the use of generative AI in marketing,” Bansal said.
Notes for 2023
Investment and growth in the artificial intelligence market at large are expected to skyrocket through the next several years, according to Fortune Business Insights, which also reports that the sector will be worth over a trillion dollars by 2029.
Research giants, including Gartner, Forrester and McKinsey, emphasize AI’s massive development. For example, Gartner analyst Afraz Jaffri recommends leaders “pay particular attention to innovations expected to hit mainstream adoption in two to five years” to gain a competitive edge. Forrester projects that AI software will grow 50% faster than the rest of the software market. And McKinsey researchers too expect AI’s adoption rate to continue to increase.
“Companies can’t let the vast power and the promise of AI/ML stop them from taking advantage of its capabilities, even as they navigate these incredibly rough economic waters,” said Lior Gavish, CTO of data observability platform Monte Carlo. “With today’s tighter budgets and leaner teams, less is truly more when it comes to optimizing the impact of AI/ML.”