Early Insights from the Monetizing Generative AI Roundtable Survey
Steven Forth is CEO of Ibbaka. Connect on LinkedIn
Ibbaka has a short survey open to inform our September 19 webinar. We are asking participants and other interested people some basic questions about generative pricing.
Some interesting patterns are emerging.
Most of the respondents believe that generative AI is transformative. Of course, this is a highly skewed data set. It is for people interested in attending a roundtable on generative AI. It does not represent the general population, or even the general technical population or business population.
Data as at Sept. 1, 2024 Morning
Still, it does suggest that among the this group there is a lot of enthusiasm.
We had two open-ended questions in the survey.
“What is an application of generative AI that worries you?”
“What is an application of generative AI that gets you excited?”
Worries
We asked Perplexity to categorize and organize the responses using Open.ai GPT4.o.
1. Security and Privacy Risks
Data Security: Concerns about data theft, data leaking to the public, and misuse of sensitive data
Cybercrime: Use of AI for security breaches, phishing schemes, and scamming through impersonation
Model Corruption: Risks of bad actors corrupting models and injecting false ideas
2. Ethical and Societal Impact
Bias and Fairness: Bias in models affecting decision-making and pricing, leading to unfair outcomes
Deepfakes and Misinformation: Creation of fake news, deep fakes, and realistic images that blur the line between real and fake, impacting public opinion and elections
Propaganda and Manipulation: Use of AI for propaganda and influencing public opinion without awareness
3. Over-Reliance and Misapplication
Over-Reliance: Dependence on AI for tasks it may not be suited for, leading to potential errors and loss of human oversight
Misapplication: Using AI in areas where it does not add value or where human oversight is crucial, such as healthcare and military applications
Generalization Issues: Overgeneralization by models and users, assuming AI can solve specific problems without proper framing
4. Economic and Competitive Concerns
Market Impact: Flattening of the playing field, making it harder to maintain a unique selling proposition
Economic Displacement: Concerns about AI replacing human jobs and the dumbing down of work
5. Technological Limitations and Hype
Model Limitations: Concerns about model collapse, slowing of model evolution, and over-hyping AI capabilities
Excessive Hype: Distraction from real issues due to over-hype and exaggerated expectations of AI
6. Ethical Use and Governance
Intent and Supervision: Worries about the intent behind AI use and the lack of supervision and responsibility
Policy and Regulation: Need for thoughtful alignment, trust, and safety strategies to prevent dangerous outcomes
Excitement
Open.ai GPT4.0 was also used to organize the responses on what gets people excited.
1. Productivity and Efficiency
Automating routine tasks and eliminating mundane work
Enhancing productivity in daily tasks and applications
Summarizing and synthesizing large amounts of information
Improving usability and functionality of apps
Saving time in research and data analysis
2. Creativity and Content Generation
Story generation and idea exploration
Video and content creation
Customizing marketing content
Generating code and visual concepts
Concept blending and creative task accessibility
3. Data Analysis and Insights
New ways to gain insights from data
Forecasting, scenario modeling, and predictive analytics
Reducing errors in data summarization and analysis
Connecting and integrating diverse data sets
Exploring counterfactuals with synthetic data
4. Knowledge Management and Integration
Capturing and amplifying human knowledge
Integrating multiple perspectives and points of view
Enhancing knowledge management practices
Connecting distributed data and knowledge
5. Collaboration and Communication
Using AI collaboratively to explore and share ideas
Facilitating communication across languages and cultures
Improving collaborative workflows and processes
6. Specific Applications and Innovations
Robotics and healthcare diagnostics
Business process augmentation and optimization
Enhancing sales and marketing workflows
Developing intellectual tools and exoskeletons
7. Personal and Professional Growth
Helping with creative thinking and brainstorming
Providing structure and guidance in tasks
Enabling more creative and focused work environments
Contact info@ibbaka.com if you would like to see the same lists categorized by other foundation models.
Impact
We also asked, “What impact will generative AI applications have on B2B software over the next 5 years?”
Again, the data is skewed by all the optimists here, but it is still suggestive.
Interesting Generative AI companies
The survey is generating an interesting list of generative AI applications. The list deserves its own analysis and post. The usual suspects are here: ChatGPT, Microsoft Copilot, Google Gemini, and Salesforce Einstein. But there is a very long tail of companies that we want to explore. Here are a handful of these applications.
Attri - generative AI and foundation models for enterprise
Covariant - generative AI for automation and robotics
DeepL - machine translation
Descript - video generation
Ivanti Neurons - unified endpoint management
Monte Carlo - end-to-end data observability
Powtoon - video and presentation generation
Phrase - machine translation
Salesloft - revenue orchestration program
SnapLogic - drag and drop data integration
StoryLab - content creation for marketing
Synthesia - text to video
And many more…
It was nice to see some people calling our Hugging Face (which now has more than 895,000 models available) and LangChain as key applications.
Will Generative AI Continue to Improve?
The capabilities of generative AI applications are generally expected to grow, but opinion was evenly divided on whether that growth will be linear or exponential.
This makes a big difference to innovation and pricing strategy. Linear growth tends to sustain innovation. Incumbent vendors can keep up and layer generative AI into existing applications. Exponential improvements are harder to manage and tend toward disruptive innovation.
How will generative AI applications be priced?
We explore this by asking about pricing metrics. The pricing metric is the unit of consumption for which the buyer pays (the value metric is the unit of consumption by which a user gets value). This will also be the subject of a future post as it needs a lot more analysis, but here are the top level results as of the morning of Sept. 1.
There is a mix of metrics here. The common generative AI metric of tokens comes up a lot (it will be interesting to see the clustering patterns to see what other metrics are used with tokens). Users also come up quite a bit, which is not surprising as applications like ChatGPT, Microsoft Copilot, Google Gemini, and Perplexity were all cited as examples of compelling generative AI applications.
Other metrics that are more associated with vertical applications include process-based and domain-specific pricing. Almost 20% called out managed services, which suggests that part of the market, probably serving enterprises, will provide customized solutions integrating multiple applications and models.
Pricing Methods
How will these applications get priced? Again there is a diversity of approaches with value based pricing leading the way. One topic people want to see covered in the roundtable is “How does one apply value based pricing to generative AI?”
It is nice to see generative pricing getting some recognition (no doubt a sign of the bias in the survey). This is an approach to pricing generative AI applications that is being developed at Ibbaka. You can learn more about generative pricing here.
Book a demo today and discover how our unique B2B SaaS solution can increase sales, improve retention, and drive growth.
