4. The speed and scale of generative AI has created an
unprecedented opportunity
4
5. 34%
of businesses lack
the necessary AI
skills, expertise
or knowledge
74%
of leaders haven’t taken
the necessary steps to
reduce bias in the
organization’s AI
1 in 5
Leaders cite difficulties
integrating data across
any cloud
Company leaders still
face challenges scaling AI
Source: 2022 AI Adoption Index: https://www.ibm.com/downloads/cas/GVAGA3JP
25%
of organizations lack
the tools or platforms
to develop models
5
6. As leaders adopt
AI, they need
to consider
three things:
1 How to create
competitive edge
2
How to scale AI
across the business
3
How to advance
trustworthy AI
6
7. Customer service App modernization
Talent
40%
Improvement in HR
productivity
70%
Contact center cases
contained by conversational
AI
30%
Productivity gain in
application modernization
• Talent acquisition
• Performance management
• Employee data management
• Employee communications
• Learning & event management
• Customer profile / demographics
• Case deflection
• Agent intent efficacy
• Agent assist
• Mobile FAQ w/ answers
• Automated code generation
• Customizable standards
• Playbook generation
• Model tuning
• Code attribution
Content Generation,
Classification
Retrieval-Augmented Generation,
Summarization, Classification
Summarization,
Content Generation
Train and tune relevant foundation
models using customer specific
datasets to improve customer
satisfaction
Automate code generation and
reduce cycle time for modernizing
applications, based on
requirements and business rules
Train and tune relevant
foundation models using
company-specific HR data
spanning hybrid environments
Generative AI tasks
Three proven, high
impact use cases
as starting points
with IBM
7
8. Ovum Medical: Trusted Experts.
Answers they agree on.
“My engagement platform is sitting on
18K clinically validated pieces of content
and thousands of hours clinically
validated video content ... Actually, I can
have internal chat experience that our
trusted content so that we can scale even
faster for the patients that we serve. It’s
really like, it solves the things that have
been on our way around for five years
being able to scale the engagement side
of the platform.”
Alice Crisci
Co-Founder and CEO of Ovum Medical
Business challenge
With over 65,000 users on Ovum
Medical’s question-and-answer
platform, the company was
searching for a solution that
could interact with patients at the
scale, while maintaining the care
and empathy of a Health Care
professional.
Solution
The Ovum AI-powered solution
addresses personal and private
fertility questions through a panel
of certified experts. It also offers
a scheduling feature, enabling
patients to book medical
appointments when further care
is advised.
Outcome
• Time saved for human patient
assist agents
• Launched 2 Virtual Agents in
2 months
• Ability to build assistants
leveraging natural language
vs. code
8
9. Make Music Count: Boost Math Skills
with Popular Piano Hits
“This was not a huge lift, there is a place
where we can start and grow to, and the
AI tool is going to learn as we use it.”
Marcus Blackwell
Founder & CEO of MMC
Business challenge
Make Music Count’s application
have engaged with 377 schools
or 60K students sign ups via their
website and 20K individual
downloads of the application in
the Android and Apple app store.
MMC needed a virtual assistant to
co-coordinate the discussions
with multiply students.
Solution
Education platform to help
improve math skills by playing
the latest popular songs on the
piano with our STEAM and SEL
curriculum and app for 2nd -
12th grade students
Outcome
• Saving on investment of
customer support headcount.
• Growth in application
capability.
• Ai engagement with students,
teachers and tutors on the
platform.
9
10. Try 6 Months of watsonx
Leverage Llama-2
& other Hugging Face models!
Sign up at Booth #222
Build with IBM
Editor's Notes
Q: HOW HAS THE MARKET SHIFTED FROM FOUNDING OAK9 TO TODAY?
We have seen dramatic growth in the market around AI.
Just look at the last five years: according to a McKinsey survey, enterprise artificial intelligence (AI) adoption has more than doubled. Clients are speaking about AI with their time, money, and interest. They know there's something here — they're experimenting — they’re establishing new practices. A new movement is happening, and IBM must be a player in that momentum.
Source:
1McKinsey Global Survey on AI, 2022: https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai-in-2022-and-a-half-decade-in-review
IDC estimates the AI Services market will nearly double from $36 billion USD in 2023 to $65 billion in 2026.
As these stats indicate, AI is going mainstream in business but with the productivity gains, come risks.
79% of executives say their organization is challenged with ensuring their AI models are responsible, secure, and free from discrimination and bias.
Our years of AI experience show us that getting to enterprise AI at scale requires a composable, multi-model strategy and a human-centric, principled approach.
Q: WHAT ARE THE CHALLENGES LEADERS ARE FACING TODAY WITH SCALING AI?
Company leaders still face challenges scaling artificial intelligence (AI). Challenges include:
Integrating data across any cloud
Access to tools or platforms to develop models
Reducing bias in the existing AI
Access to skills, expertise or knowledge
Note:
Bias is a type of error that can occur if an artificial intelligence (AI) models’ output is skewed by the model’s training data. For example, a model may associate specific traits or professions with a certain race or gender, leading to inaccurate predictions and offensive responses.
Source:
2022 AI Adoption Index: https://www.ibm.com/downloads/cas/GVAGA3JP
Q: HOW SHOULD COMPANIES, AND PARTICULARLY STARTUPS EVALUATE AI OPPORTUNITIES?
How to create competitive edge. The key to a client’s success is in the core of their business—the essential activities and capabilities that are fundamental to who you are and who you serve. Whether based on machine learning (ML) or foundation models, the more customized a client’s artificial intelligence (AI) models are to those priorities, the better they will be able to serve their customers and deliver real business value. Foundation models make it possible to fine-tune AI to an enterprise’s unique data and domain knowledge with a specificity that was previously impossible (or at minimum too disrupting because of the data labeling requirements), so it is essential to let the business strategy guide the data strategy. To be truly impactful, AI should integrate into existing workflows and systems, automating key processes across areas such as customer service, supply chain, and cybersecurity.
How to scale AI across the business. AI is only as good as the data that fuels it and it’s critical to identify the right data sets from the beginning: poor quality data can cause projects to falter, while businesses cite excessive data complexity and challenges with integration as posing major obstacles to AI adoption. Clients should ask themselves: “What is the most critical data?” and “Which data provides the strongest competitive advantage?” Also, the data that fuels business processes is often widely distributed, so businesses must create AI-ready architectures. Data, in fact, is everywhere: in on-premises data centers, on mainframes (IBM Z), in private clouds, public clouds, and on the edge. In order to successfully scale AI efforts, clients need the ability to make use of all their data, wherever it resides. A hybrid cloud architecture provides the data foundation for extending AI deep into the business.
How to advance trustworthy AI. If a business is depended upon to provide essential services, or quickly deliver accurate information, insights, or recommendations at scale, their systems need to help maximize availability and minimize errors. If the models contain bias, or if AI models are misleading, “hallucinate,” or are not explainable, the risk and cost of reputational damage and regulatory fines could be high. AI must be explainable, fair, robust, transparent, and prioritize and safeguard consumers’ privacy and data rights to engender trust. Data and AI lifecycle management is an important part of improving data access, applying governance, cutting costs, and getting quality models into production faster.
Notes:
Bias is a type of error that can occur if an artificial intelligence (AI) models’ output is skewed by the model’s training data. For example, a model may associate specific traits or professions with a certain race or gender, leading to inaccurate predictions and offensive responses.
Hallucination is a well-known phenomenon in large language models (LLMs) in which the system provides an answer that is factually incorrect, irrelevant, or nonsensical because of limitations in its training data and architecture; more concerning is the hallucinated answer sounds plausible.
A large language model (LLM) is a type of machine learning model that has been trained on large quantities of unlabeled text using self-supervised learning and can perform a variety of natural language processing (NLP) tasks (even when that language is a programming language). Output may range from books, articles, social media posts, online conversations, and even code. The architecture of an LLM consists of layers of neural networks that learn to generate language in a way that is similar to how humans use language
Machine learning (ML) refers to a broad set of techniques to train a computer to learn from its inputs, using existing data, and one or more “training” methods, instead of being explicitly programmed. ML helps a computer to achieve AI.
Natural language processing (NLP) is the technology that gives computers the ability to understand text and spoken words in much the same way human beings can. NLP combines computational linguistics — rule-based modeling of human language — with statistical, machine learning (ML), and deep learning (DL) models. These technologies enable computers to process human language in the form of text or voice data and to ‘understand’ its full meaning, complete with the speaker or writer’s intent and sentiment.
Deep learning (DL) is a technique for implementing machine learning (ML) that relies on deep artificial neural networks to perform complex tasks such as image recognition, object detection, and natural language processing (NLP). Neural networks are a set of algorithms, modeled loosely after the neural networks found in the human brain, which are designed to recognize hidden patterns in data. The “deep” in DL refers to a neural network comprised of more than three layers (which include the input and the output layers).
Neural networks are a set of algorithms, modeled loosely after the neural networks found in the human brain, that are designed to recognize hidden patterns in data.
An algorithm is a procedure used for solving a problem or performing a computation. Algorithms act as an exact list of instructions that conduct a sequence of specified actions in either hardware- or software-based routines.
Sources:
IBM Global AI Adoption Index 2022: https://www.ibm.com/downloads/cas/GVAGA3JP
IBM Institute for Business Value | Research Insights, A comparative look at enterprise cloud strategy, September 2022: https://www.ibm.com/downloads/cas/MWEMP0NJ
Q: WHAT COMMON USE CASES CAN AI SOLVE FOR STARTUPS?
The best opportunities to derive benefit from AI in business today include Digital Labor, IT Automation, Security, Sustainability, and Application Modernization.
Resist the urge to indulge in what fantastical things artificial intelligence (AI) can do. Organizations that stay grounded in the practical things AI can help them achieve – both short and long term – are more likely to succeed.
According to Gartner, business leadership tends to overestimate the impact of AI and underestimate its complexity – requiring data and analytics leaders to manage the business’s expectations, or risk costly project failures.
You can reduce this risk by identifying your key priorities and leveraging technology like IBM’s to deliver results
Source:
Gartner Research: https://www.gartner.com/en/doc/730970-what-is-artificial-intelligence-seeing-through-the-hype-and-focusing-on-business-value
TELL US ABOUT HOW IBM HAS HELPED OTHER COMPANIES ACHIEVE SUCCESS IN THEIR AI STRATEGY
Go beyond purchasing AI and develop you AI Strategy and Partnership. With IBM, we have partnered with companies such as Dubber and Krista to co-create successful, enterprise platforms.
Raj to speak about 1-2 of these
Ovum Medical is an example of how embedding IBM’s Watson Assistant enables our partners to build accurate and empathetic interactions with patients at scale: 24X7 and 365 days a year.
TELL US ABOUT HOW IBM HAS HELPED OTHER COMPANIES ACHIEVE SUCCESS IN THEIR AI STRATEGY
Go beyond purchasing AI and develop you AI Strategy and Partnership. With IBM, we have partnered with companies such as Dubber and Krista to co-create successful, enterprise platforms.
Raj to speak about 1-2 of these
Make Music Count is a math curriculum in an app taught by learning how to play your favorite song on the piano. Being centered on engagement, IBM Watson Assistant enables MMC to assist users in enhancing engagement with users over time by learning and training via integration of IBM’s embeddable AI software solution..