At the recent annual Think conference, IBM announced the release of its new AI and data platform, IBM watsonx. The platform aims to help companies scale and accelerate AI with trusted data, offering everything from hardware to data storage to ML and AI foundation models. IBM’s new AI product is a direct response to other products like OpenAI’s chatbot ChatGPT, which has dominated public awareness since December.
Everything may be a little too late for IBM to take the lead in the market.
They’re not the first to announce a platform of this kind,” Omdia chief analyst Bradley Shimmin said in an interview with Data Center Knowledge. “Almost every major platform with AI development tools, such as Google and AWS’s Databricks, is already working on this. I don’t think he’s in a position to show leadership unless he surprises us with it.”
What are the components of IBM’s watsonx platform?
The watsonx platform contains two parts: watsonx.ai and watsonx.data. Watsonx.ai is an AI studio that combines the power of IBM Watson Studio with modern generative AI capabilities that harness the power of underlying models. IBM’s watsonx.data is a fit-for-purpose data store built on an open lakehouse architecture optimized for governed data and AI workloads.
Expected to be generally available by July 2023, it can manage workloads in both on-premises and multi-cloud environments. The solution also offers built-in governance tools, automation, and integration with your organization’s existing databases and tools, simplifying setup and user experience.
To transform customers from just “users”, IBM also creates a governance toolkit as part of the watsonx product.
“They may be able to take advantage of AI,” said IBM chairman and CEO Arvind Krishna in a statement. become a target,” he said. IBM watsonx enables customers to rapidly train and deploy custom AI capabilities across their business while maintaining complete control over their data. ”
Additionally, the company promises to reduce data warehouse costs by up to 50%.
Shifting focus from giant AI models to effective solutions
IBM’s watsonx platform aims to be a leader in AI, but OpenAI CEO Sam Altman thinks the era of giant AI models may be coming to an end. At an MIT event in April, Altman declared that future advances in AI will require new ideas, not just larger models.
“I think we’re at the end of the era of giant giant models,” Altman told the audience. “We will improve in other ways.”
However, it’s worth noting that some of the AI innovations Altman mentions may involve watsonx. watsonx uses machine learning and AI-founded models to enable enterprises to rapidly train and deploy custom AI capabilities, potentially leading to new breakthroughs in this space.
Tarun Chopra, vice president of product management, data and AI at IBM, said, “ChatGPT can write jokes, pictures, and even songs and entire movie scripts, but this kind of consumer AI output is something that companies can’t afford. It is not required,” he said in an email. “Companies just beginning their AI journey need more than ever to seamlessly adopt the technology.”
According to Chopra, businesses need absolute confidence that the AI used for mission-critical decisions and outputs can be trusted while leveraging their own data.
Additionally, “watsonx allows companies to own their data on their own infrastructure,” Chopra said.
Shimmin said he does not intend to compare the two platforms because OpenAI is building a closed system with a number of new properties that allow users to run different modalities in one model. I’m here.
“So what IBM is talking about is helping companies get the same type of functionality they get from the GPT model, but in a way they are more familiar with and more comfortable with,” Shimmin explained. .
By taking existing machine learning algorithms and scaling them up to sizes never imagined before, OpenAI was able to extend the suite of advanced AI for consumers. The latest of these projects, GPT-4, was probably trained using trillions of words of text and thousands of powerful computer chips. The product’s rapid deployment has sparked an AI arms race in the tech industry and raised alarm among some AI ethicists and civil servants, who fear the technology could spread misinformation, replace jobs, or otherwise We are concerned about the possibility of causing serious damage to users in the form of
This is what Watsonks eliminates additional Chopra, as IBM focuses on “meticulously managing everything that goes into the model.” A basic model was developed to classify hate, profanity, license restrictions, and bigotry data.
But Shimmin said Speaker OpenAI is working to help companies bring their own data in a secure way. Shimmin said Salesforce is partnering with his OpenAI to build a “walled garden” of its own iteration of his OpenAI to control and secure data access.
IBM Watson AI, the class of over-promising and under-delivering
Watson rose to fame on the popular network game show Jeopardy when he faced two of the most successful contestants, Ken Jennings and Brad Rutten. As expected, Watson crushed his human competitor, winning $77,147, while Jennings and Rutter took home $24,000 and $21,600, respectively.
Watson’s fame was short-lived, being seen as a one-trick pony and a cautionary tale of technology’s overpromises and underperformance. However, the newly revitalized AI trends facilitated by OpenAI’s ChatGPT-4 may spark renewed interest in the public’s perception of Watson AI. This could make up for the rise and fall of Watson AI in 2011, which left concerns about IBM’s commitment to emerging technologies. One of his reasons for this is that IBM is the cloud’s biggest competitor for his infrastructure, Amazon his web his services and his Microsoft Azure data centers with his Watson and its array of his AI tools. made it possible to do it. Operating software to train machine learning models is less expensive and more efficient when run in the same place.
Moreover, earlier Reuters reports said IBM’s original Watson didn’t take off because it was too costly for companies to deploy. IBM is currently marketing its business-focused watsonx as an AI development platform that businesses can use to build their own models for a variety of purposes, including customer care and writing code.
Simin said there is a race to innovate on smaller metabase models, such as Facebook’s LLaMA-based model that drives the enterprise.
Shimmin said it’s not just a few super-large language models like GPT or PaLM, it’s small models that are tuned using enterprise data, and they’re secure, governed, governed, and transparent. He said that it is a fine adjustment to be made in a difficult environment.
“And that’s what IBM is trying to establish here with watsonx,” says Shimmin. “So I applaud their approach.”
Will watsonx lead the enterprise AI solutions race?
As the AI industry continues to expand rapidly, IBM’s one-stop-shop approach to your AI and server needs may give you an edge over your competitors. While investments in AI continue to draw buzz about global changes akin to the Industrial Revolution, given that IBM was reported earlier this month, many, including IBM employees, are skeptical about the company’s upstart. Concerned about approaching technology. As many as 7,800 jobs are at risk after a hiring pause in favor of AI.
In a Q&A ahead of IBM’s Think conference, Krishna predicted that Company A would take over “repetitive back-office processes” from human employees.
“With this, we find that 30-50% of that volume of tasks can be easily performed, and we are able to perform the tasks with a level of proficiency equal to or better than that of humans,” said Krishna.
“Many of them will be accepted immediately from this year and fully come to fruition in the next three to five years,” Krishna added.
