The Health Strategist
research and strategy institute
for continuous transformation, of
in person health and digital health
Joaquim Cardoso MSc
Chief Research and Strategy Officer (CRSO)
9 de maio de 2023
ONE PAGE SUMMARY
IBM has announced the launch of a generative artificial intelligence (AI) platform called “watsonx,” aimed at helping businesses build AI-based applications and manage their AI training data.
- The platform offers businesses greater control over their data, allowing them to run it on-premises or on public clouds in a more private instance.
- IBM’s platform includes a governance component to manage fairness and ethical use of data and AI.
- The AI platform will be available on IBM Cloud in July, with support for on-premises and other cloud platforms coming later.
- IBM is also providing consulting services to assist customers in adopting the generative AI platform and has partnered with machine-learning startup Hugging Face to make open-source AI models available.
IBM’s move aligns with other tech giants like Microsoft, Amazon, and Google, who have recently announced their own generative AI platforms and developer services.
- Generative AI presents an opportunity for these companies to boost their cloud-computing businesses by enabling companies to build applications and automations at lower costs and with greater ease.
- IBM’s differentiating factor is its platform’s compatibility across cloud providers and data centers, accommodating businesses running their IT infrastructure on-premises.
IBM’s focus on trust and data privacy is evident in its training of foundation models on curated and vetted enterprise data with auditable data lineage.
- The company recognizes that trust and transparency concerns are obstacles to wider adoption of generative AI and aims to address these issues.
- Business leaders are increasingly seeking ways to experiment with generative AI while ensuring data privacy and implementing safeguards.
IBM believes that combining businesses’ proprietary data and industry expertise with large language models will make generative AI a reality for enterprises in the near term.
- The company emphasizes that its approach is not focused on using generative AI for customer support in a Shakespearean manner, but rather aims to provide businesses with a practical and tailored solution.
Overall, IBM’s entry into the generative AI platform market strengthens its position as a provider of AI solutions for businesses.
- By offering control, governance, and compatibility across various infrastructure options, IBM aims to cater to the diverse needs of enterprises while addressing concerns around trust, transparency, and data privacy.
DEEP DIVE
WSJ: IBM Announces Generative AI Platform Aimed at the Enterprise
The Wall Street Journal
By Belle Lin
May 9, 2023
IBM on Tuesday unveiled a service that helps businesses build generative artificial intelligence-based applications and manage their AI training data-joining the ranks of technology giants seeking to become the preferred platform for companies to build on the AI behind ChatGPT.
“It lets everybody tap into the potential for large language models and generative AI,” said Chief Executive Arvind Krishna. “But some want to have much more control of the data. They want to run it possibly on-premise, or they possibly want to run it on a public cloud but in a more private instance-it’s for people who want to do that.”
IBM’s platform, called “watsonx,” also includes a governance component for businesses to manage “fairness” and “ethical use of both the data and the AI,” Mr. Krishna said. The AI platform will be available in July on IBM Cloud, and on-premises and other cloud platforms later on, according to IBM. The governance tools will be offered later this year.
Tuesday’s announcement represents one of IBM’s first steps in making generative AI capabilities more widely available to its customers. While the company has long invested in AI-including the previous iteration of its Watson AI system-its new chapter is strictly focused on business applications for the technology, Mr. Krishna has said.
IBM also announced a new consulting team to help customers start using its generative AI platform, and a partnership with machine-learning startup Hugging Face to make open-source AI models available. In doing so, the company is betting that businesses will need a provider like IBM to “bring that open-source innovation in a way that enterprises can consume,” said Darío Gil, senior vice president and director of IBM Research.
Tech giants Microsoft, Amazon.com and Google, all recently announced platforms and developer services that help companies use generative AI to build applications and automations at lower cost, and with greater ease, than if they were to do so on their own. For each company, generative AI is an opportunity to drive greater use of their respective cloud-computing businesses -the backbone for running and developing AI-powered apps.
IBM’s generative AI platform works across cloud providers and in data centers, the company said, helping to distinguish its offering as available to companies that run their information-technology on-premises, not just in the cloud. “We allow an enterprise to use their own code to adapt the model to how they want to run their playbooks and their code, and then they can deploy for themselves without any danger of their code leaking,” Mr. Krishna said.
As part of its assurances to enterprises regarding data privacy, IBM said its own foundation models are trained on “a large, curated and vetted set of enterprise data” with “auditable data lineage.” The company’s recent survey of business leaders found that 30% said trust and transparency issues are holding them back from adopting generative AI, and 42% cited data privacy concerns.
Instead of banning ChatGPT outright, business technology chiefs are increasingly looking for ways to experiment with its capabilities and apply it with guardrails and data transparency.
The combination of businesses’ own data and industry expertise, plus large language models, will make generative AI a near-term reality for enterprises, said Rob Thomas, IBM senior vice president for software and chief commercial officer. “We’re not trying to do customer support in the style of Shakespeare, to use the popular consumer use case that you see. It’s a different approach here,” he said.
Originally published at https://cio.cmail19.com.
Names mentioned
- Arvind Krishna: Chief Executive of IBM.
- Darío Gil: Senior Vice President and Director of IBM Research.
- Rob Thomas: IBM Senior Vice President for Software and Chief Commercial Officer
