Industry executives and experts share their predictions for 2023. Read them in this 15th annual VMblog.com series exclusive.
10 Language AI predictions for 2023
By Amit Ben, founder and CEO of One AI
Natural Language Processing (NLP) has come a
long way in the past two decades, with numerous professionals contributing to
its development and the creation of products serving hundreds of millions of
users. In 2022, NLP finally reached mainstream recognition and utilization. The
field has exciting potential for future growth and development. As we look
towards 2023, here are some predictions for the future of Language AI:
1. The Bar will be Set and Raised - we predict a shift in the way humans interact with machines, as
people will come to expect products and services with language AI
capabilities. In the race to embrace AI-driven capabilities, those that
fail to keep up and leave humans isolated from AI creation and moderation
superpowers, will be left behind. NLP will be used in various industries
such as sales, customer success, health, and law.
2. Productization of NLP - we
expect to see the productization of NLP, similar to what has happened to
other fundamental technologies with companies such as Twilio and AWS.
Rather than building their own solutions or using open source models,
product companies will turn to platforms and APIs to provide them with the
NLP capabilities they need. This trend has already begun with transcription
and translation, and we expect to see it extend to other capabilities as
well. By leveraging these platforms and APIs, companies can save time and
resources while still gaining access to advanced NLP technology.
3. "We Have Two Ears and One Mouth", as Epictetus said. With the growing success of language
generation models and tools like ChatGPT, NLP models will continue to
improve their ability to generate coherent, natural language text. These
models will be utilized in various applications, such as chatbots,
language translation, and content creation. However, we predict that
Natural Language Understanding (NLU) will become increasingly crucial. NLU
allows businesses to better understand and respond to the needs and
preferences of their customers, improve customer experience and increase
satisfaction and loyalty. It can also improve the efficiency and accuracy
of business processes by automating tasks that require language
comprehension, such as customer service inquiries or data entry.
4. Factual Accuracy - as NLP
applications become more widespread, users and businesses will demand a
greater focus on factual correctness and proper referencing of sources.
While grammar and form are important, ensuring that the information being
provided is accurate and can be traced back to a reliable source will
become increasingly crucial. This will require a significant effort from
developers and researchers to prioritize factual correctness in their NLP
models and systems.
5. Language Analytics in Scale
- the emergence of large-scale language analytics is revolutionizing the
way we process and analyze language data. These tools allow us to quickly
and accurately digest massive amounts of language, including social media
content, customer relationship management (CRM) data, legal documents, and
electronic health records. This is enabling organizations to gain new
insights and make more informed decisions based on a deeper understanding
of language-based data.
6. The Other 95% - as the
demand grows globally, there will be a need for multilingual NLP solutions
to meet the needs of non-English speaking populations. This shift towards
multilingual NLP will enable organizations to communicate more effectively
with diverse audiences.
Currently, most
investment has been in English, with Chinese following closely behind. However,
NLP is more dependent on culture and language than any other technology, and
the inherent differences between languages and cultures will lead to English
and Chinese NLP technologies evolving in fundamentally different ways.
7. AI-aided Software Development - with the popularity of code-generation tools like Github
Copilot, there will be an increase in the use of Language AI for code
generation and analysis. In the future, we can anticipate the emergence of
new tools that encompass more parts of the software development cycle,
including architecture planning, code analysis, test generation, and even
complete app coding.
8. Open-source Will Stay Open (for Now): Microsoft, and perhaps other tech giants, will attempt to
acquire HuggingFace, the leading provider of open-source AI models, but
the deal will ultimately fail. HuggingFace will remain an independent
company and continue to provide the largest library of open-source AI
models to users around the world. This will be seen as a victory for the
open-source community and a testament to the importance of keeping such
resources freely available to all.
9. You are What You Curate? -
as AI becomes more advanced and widespread, it is likely that humans will
increasingly focus on curation rather than creation. Curation involves
selecting, organizing, and presenting information in a way that adds value
to the user, while creation involves creating new content or ideas. As
humans become more adept at using NLP to help them curate and organize
information, they may be able to more effectively disseminate knowledge
and ideas, and make more informed decisions.
10. "Big Brother is Filtering (for) You" - Human beings have always consumed content through various
filters, such as editors in newspapers or algorithms on social media
platforms. The next step in this process is the emergence of language
AI-powered filters, which will set the way we discover and consume
content. Think of an attorney using AI to sift through the records of a years-long
legal case to help write the closing arguments, or an individual asking AI
to summarize their work before sending in their weekly report.
In 2023, as the AI industry landscape map is getting crowded, we can
expect significant progress in NLP with a focus on product development,
real-world adoption, multilingual capabilities, and the emergence of new
applications. This will lead to a wave of innovation that is currently
unforeseen. Just as the iPhone inspired various innovative applications, such
as DoorDash, TikTok, or digital wallets, NLP technology has the potential to
drive the creation of applications that will greatly impact our lives. Keep an
eye out!
##
ABOUT
THE AUTHOR
Amit Ben is the Founder and CEO of One AI. He
has been fascinated by language technologies since he was a child and has
dedicated his career to advancing the field. Amit's passion for NLP led him to
co-found NanoRep, a company that he later sold to LogMeIn. At LogMeIn, Amit led
the AI lab and developed the technology that powers products enjoyed by
hundreds of millions of users, such as GoToMeeting, Bold360, and LastPass.
After working on language AI capabilities for products for the past two
decades, he realized that the technology had reached the point where it could
be offered as a service, and decided to build it one last time, as a service
that any company could use.