At Innovate, SAS relies fully on (generative) AI, but in its own way. In times when everything revolves around productivity and performance, a third component is paramount for SAS: trust.

SAS Innovate is the annual high mass for everyone who cares about the data specialist. None other than CEO and co-founder Jim Goodnight will make a surprise appearance on opening day. The now 81-year-old Goodnight no longer romps around the stage like a young foal, but the man has not yet lost his sharp mind and sense of humor.

The recognizable melodies can be heard at times during the keynotes. “Generative AI is the next wave in the way we interact with systems and data. It has its origins in a historical context: first we developed data lakes to store data, then databases to organize data, and now GenAI to naturally interact with data,” said Bryan Harris, CTO of SAS .

Still, Harris doesn’t want to fall too far into clichés. “The intention should not be to simply sell hype that people will then be disappointed with. Don’t invest in generative AI just because the big players tell you to. AI in itself is not the solution to your business problems, but rather a means to achieve that solution. If you ask the AI ​​bad questions, you will get bad answers. People play a central role in how and why we implement technology.”

Golden time for developers

The SAS Gospel has three parts: Productivity, Achievement and Confidence. Productivity is the focus on the first day of the conference. SAS goes full throttle and announcements follow at a rapid pace, from one day onwards Data Creator to a dedicated co-pilot for Viya, the glue of the SAS ecosystem.

These days, no platform seems complete without a built-in AI assistant. What makes SAS Copilot different from the other AI assistants with a similar name? Harris doesn’t avoid the question: “What sets us apart is that we can be cloud agnostic.” Our solutions provide a consistent experience no matter where customers deploy them. In addition, respect for the integrity of company data is embedded in all our products: in this respect, Copilot is an extension of what we already offer with Viya.”

“There’s no better time to be a developer than now,” he says Strategic AI consultant Marinela Profi is a bold statement. “Artificial intelligence has the potential to increase our most important quality: creativity. When you make developers more productive, you make your entire company more productive.” Jensen Huang, CEO of Nvidia, may be feeling personal since he previously claimed that programming was no longer a valuable skill.

Since SAS is a data company through and through, the importance of data for AI is also highlighted. Without qualitative data, there is no AI, Profi makes it clear. “With every major technological development, we always come back to data.”

AI is not the solution to your business problems, but a means to achieve that solution. If you ask the AI ​​bad questions, you will get bad answers.

Bryan Harris, CTO SAS

Bigger is not always better

Day two of SAS Innovate focuses on performance. To achieve this, SAS pays attention not only to speed, but also to scalability and cost of implementation. The company is therefore following the current development of LLMs, which are becoming increasingly larger and heavier, with caution. This moves AI models further away from actual problems that companies face.

Harris explains this perspective: “Practical business problems cannot be solved with an LLM. In concrete practical situations, a “lightweight” model that focuses on the specific problem is more useful. When we talk about lightweight AI, it’s not just about the complexity and number of parameters of models, but also the speed and cost of implementing them in the IT environment.”

SAS’ answer is “models”: lightweight, industry-specific models that companies can purchase as pre-built “packages.” Udo Sglavo, Vice President of Advanced AnalyticsHe gets the honor of presenting it: “This is a new way to bring algorithms to the masses. There are developers of AI solutions, but many companies just want to be able to use AI.”

Many companies simply want to use AI.

Udo Sglavo, Vice President of Advanced Analytics SAS

Trust first

“Productivity and performance are worthless if you don’t get reliable results,” notes Jared Peterson (VP Platform Engineering) At. This builds a bridge to the third main theme of SAS Innovate: trust. The fact that this topic is discussed last does not mean that it is at the bottom of the list for SAS. On the contrary, the company wants to make a name for itself in the tech industry with its focus on “honest” AI that has nothing to hide from the outside world.

Reggie Townsend Vice President of Data Ethics, leads this mission within SAS. While there was room for a humorous note in previous lectures, his speech was bitterly serious. Townsend silences the audience several times. “There is a trust gap around AI: people are rightly wondering how AI will impact their security. This fear is increasing rapidly. Every technological breakthrough has negative side effects. We can choose to let it continue or do something about it.”

Who then bears responsibility for making AI reliable is a complex question for Townsend. “I think you have to be careful who you point the finger at. AI creates a networked system in which everyone has some responsibility. Sociotechnological problems require collective action. To fight for what’s right, you have to be willing to get in the ring and accept setbacks.”

To fight for what is right, you must be willing to step into the ring and face adversity.

Reggie Townsend, Vice President of Data Ethics SAS

During Innovate, SAS encourages you to not only think in terms of productivity and performance, but also to build trust in the technology before the first line of code is written. The company hopes not to spread that message in the desert, even though it’s not far away in Las Vegas.