Home / Security / Businesses need pricing clarity as generative AI services hit the market

Businesses need pricing clarity as generative AI services hit the market

Dollar sign on abstract AI background

Aleksandra Malysheva/Getty Images

There needs to be clarity around how exactly generative artificial intelligence (AI) will be charged, as market players rush to push out their offerings and businesses look to avoid bill shock. 

Transparency around the usage and commercial model is something organizations are asking for, so they can avoid escalating hidden costs, said Tim Dillon, founder and director of Tech Research Asia. There is general concern they will experience bill shock from a consumption-based model, similar to how some had to deal with this challenge during the early days of cloud, he noted.

Also: We’re not ready for the impact of generative AI on elections

It is something vendors such as Salesforce will have to figure out as they ramp up their generative AI service offerings, said Dillon in an interview with ZDNET, on the sidelines of Dreamforce 2023 held in San Francisco this week. 

The adoption of these tools can grow organically within an organization and, hence, can lead to a lack of control and awareness of their consumption. There also often are no policies guiding the use of generative AI, he said, adding that research suggests 40% of organizations in Asia-Pacific Japan have informal policies around such tools, while 60% have formal policies in place. 

Concerns around bill shock are compounded by the softening economy, with companies in the region facing potential budget cuts, he noted. And if prices are tagged in US dollars, consuming generative AI can be an expensive proposition for businesses in some Asia-Pacific markets. 

Also: The moment I realized ChatGPT was a game-changer for my business

Acknowledging that concerns about bill shock were valid, Gavin Barfield, Salesforce’s Asean vice president and CTO of solutions, said pricing models still are being defined as generative AI services gradually are rolled out. 

“We’re in the early stages, so all companies are wrangling with these issues,” Barfield told ZDNET. He noted that the same issues had surfaced when cloud services were first launched. 

“As the market and product mature, these things will get ironed out,” he said, adding that market players will need to find ways to price generative AI services. Salesforce itself is looking at a variety of pricing models but for now has opted for a credits-based system for a couple of services, according to Barfield. How much credits are consumed depends on how the AI model is called to run the query. 

In July, Salesforce announced that Sales GPT, which is shipped with Sales Cloud Einstein, is available at $50 per user per month and includes a limited number of Einstein GPT credits. Service GPT, shipped with Service Cloud Einstein, also is priced at $50 per user per month and includes a limited number of Einstein GPT credits.

Also: How trusted generative AI can improve the customer experience

Customers of either generative AI services can purchase Enterprise Expansion packs for more credits when their usage grows. 

Because generative AI services are based on a usage model, it is critical that companies can monitor their consumption, said Jan Morgenthal, chief digital officer of Singapore telco, M1. 

Speaking to ZDNET at the conference, he noted the need to be able to measure and forecast how much these tools are used within his organization. M1 currently uses several AI tools from various vendors, including Salesforce, and is also testing generative AI services. 

Having a dollar a value, for instance, will enable him to manage the number of queries that should be made with these tools.

Morgenthal noted that, depending on the complexity of a particular use case and the AI model needed to automate or generate a response, it may not make sense in terms of the ROI (returns on investment) to power the query with generative AI. 

This is an issue that companies will need to be cautious about or costs can escalate. The automation gained from the generative AI then may not be worth the cost of its delivery, he said. 

Also: Everyone wants responsible AI but few people are doing anything about it 

It also means organizations have to map out the processes, including data availability, needed to run a query and achieve the desired outcome, so they can measure the cost of applying generative AI to the use case.

Letting customers create their own prompts

Salesforce this week previewed new generative AI offerings that its executives said would enable enterprise customers to more easily customize these tools to support their operations. 

Among them is the Einstein Copilot, touted as a conversational AI assistant that can be integrated with any Salesforce application, enabling users to ask questions in natural language.

Responses are generated based on proprietary company data powered by Salesforce Data Cloud, previously called Genie. The data engine pulls together any datasets, including customer data, telemetry data, and Slack conversations, to create a unified view of the customer. 

Data Cloud currently processes 30 trillion transactions per month and connects 100 billion records daily, according to Salesforce. The data engine is now natively integrated with Einstein 1 Platform, enabling businesses to apply AI, automation, and analytics to every customer experience. 

It enables Einstein Copilot to provide options for additional actions beyond the user’s query, such as a recommended action plan after a sales call.

Also: 4 ways to increase the usability of AI, according to industry experts

Organizations that want to build generative AI applications with their own customized prompts, skills, and AI models also can do so via the Einstein Copilot Studio. It encompasses the Prompt Builder, which lets users create and test generative AI prompts that are aligned with their corporate brand and communication style. And they can do so without any technical expertise, enabling marketing executives to ask Prompt Builder to generate a tailored message based on a customer’s purchase or order history. 

Einstein Copilot Studio also includes Skills Builder, which allows companies to create custom AI-driven actions to run specific tasks. For example, it can create a “competitor analysis” skill that analyzes current market data, sales figures, and send API calls to extract data from external sources.

In addition, a Model Builder component enables organizations that want to use their own AI models. They can choose one of Salesforce’s proprietary LLMs (large language models) or integrate their preferred predictive and generative partner AI models. They can train these on data in Data Cloud without moving or copying data. 

This means Einstein Copilot can provide more accurate insights and content that are tailored to the company’s employee or customer dynamics. 

Model Builder eventually will support external LLMs that include Amazon Bedrock, Google Cloud’s Vertex AI, Anthropic, and Cohere. For now, it only supports OpenAI.W

Also: What technology analysts are saying about the future of generative AI 

Einstein Copilot currently is in pilot, while Copilot Studio will enter pilot later this fall. Einstein Trust Layer enhancements will be generally available in the vendor’s Einstein platform from October 2023. No pricing details are available for the new offerings. 

Data Cloud currently is included for Enterprise Edition or above customers at no cost. It encompasses capabilities that allow organizations to unify 10,000 customer profiles and includes two Tableau Creator licenses.

In addition, a new Einstein Trust Layer will underpin all Einstein products, providing a secure AI architecture that Salesforce said will ensure its customers’ generative AI responses are powered by quality data that are checked against potential bias and security and privacy standards. 

Integrated with Salesforce Data Cloud, the Trust Layer mitigates such risks, checking data against toxicity and brand risks, masking personal identifiable information, and not retaining customer data. 

 Einstein Trust Layer enhancements will be generally available in October 2023, according to Salesforce. 

“The reality is every company will undergo an AI transformation to increase productivity, drive efficiency, and deliver incredible customer and employee experiences,” said Marc Benioff, Chair and CEO, Salesforce. “With Einstein Copilot and Data Cloud we’re making it easy to create powerful AI assistants and infuse trusted AI into the flow of work across every job, business, and industry. In this new world, everyone can now be an Einstein.”

Based in Singapore, Eileen Yu reported for ZDNET from Dreamforce 2023 in San Francisco, USA, on the invitation of Salesforce.com.


Source link

About admin

Check Also

Gen AI training costs soar yet risks are poorly measured, says Stanford AI report

The number of significant new AI models coming out of industry has surged in recent ...

Leave a Reply

Your email address will not be published. Required fields are marked *