Main content

Salesforce kicks off AI center and training for the other 80%

George Lawton Profile picture for user George Lawton June 21, 2024
Summary:
Salesforce opens AI center and launches the AI Now Tour to bring AI to the other 80% of users. It starts with collaboration, data, change management, and learning how to ask the right questions.

Salesforce

Salesforce has cut the ribbon on its first AI center in London, which promises to make AI accessible to less technical users. The event coincided with the AI Now world tour launch in the UK, which continues with free training around the globe. Surprisingly, the training starts with learning better ways to work with data and ask questions (prompts) before moving on to learning how to use AI to perform tasks.

I was half expecting the new AI center to showcase a shiny new supercomputer with flashing lights. Instead, we were greeted with an array of two-seat desks, dual monitors, and empty space for laptops and notebooks. Attendees could wander the entire 40,000 square-foot floor with stunning views on all sides without passing through a single door. The implication was that learning to work with AI was more about collaboration to solve practical problems than solo explorations of the shiny new thing.

The AI center and the training program are part of Salesforce’s $4 billion commitment to grow the current $21 billion UK AI market to $1 trillion by 2035. Salesforce's unique approach prioritizes people, fostering collaboration with UK institutions and building trust, rather than simply focusing on AI model development.

Salesforce UK & Ireland CTO & SVP Solution Engineering Paul O’Sullivan says the majority of UK workers feel under-equipped to apply AI effectively and safely, which is a huge risk:

I think this is an opportunity for us to start closing that gap and taking people on a journey of how to apply AI to things securely… I personally believe that this is going to enable us, the UK, to really prosper. So, it's a really exciting time, but we've got to get this right, and this is why the AI hub center is going to be big for our Salesforce customers.

Closing the knowledge gap

One big opportunity lies in closing the knowledge gaps between consumer AIs like ChatGPT and enterprise AI. Consumer AI looks powerful, but it’s important to understand the risks of using these tools in the enterprise. Another opportunity lies in navigating the change management issues required for enterprises to adopt the technology responsibly.

O’Sullivan says their experience in collaborating with customers through the AI center will help refine the Salesforce Trailhead training program to help less technical users navigate change management and cultural aspects of adopting AI:

I think one of the key things is the change management element in the workforce in the future is going to be really important. As the technology starts to enable people, how do we make sure that people put it to use in a consistent way that equally supports the right levels of productivity and efficiency within that organization without potentially causing a tremendous amount of costs?

Another opportunity lies in training the next generation of users. O’Sullivan explains:

We've got plans to not only just bring this to our customers, industry experts and partners, but we also want to broaden that out to universities and high schools, and to make sure that we are building the generation the future and equipping them with the skills and the knowledge they require.

Prioritizing long-term opportunities

O’Sullivan notes it’s also important to help business users understand where AI can help create long-term value for the business and society rather than solve the wrong problem. Early in his career, O’Sullivan worked for a couple of companies ahead of the curve with amazing technology that failed to align with practical economics. One ended up helping to modernize the back-end infrastructure for most UK banks. However, its focus on growing customers came at the expense of ongoing profitability.

Today, he sees the same thing happening with enthusiasm for AI. It’s a great conversation starter, but enterprises must get the basics right first. He explains:

Right now, eight out of ten conversations start with AI, but following the conversation, ten out of ten land around ‘I need to take action on my data.’ I think it's really interesting because there’s a balance in every technology decision. There's a tradeoff. We need to be really careful, and we've got to help guide and navigate our customers so that they don't make decisions for speed and have those huge consequences associated with the impact afterward.

ChatGPT and the consumer AIs are great. There's a user interface and a model that underpins it, but there's a whole heap of public data, and this public data is just an open pool. If I chuck my enterprise data into that space, the model learns from that data and retains some of that information, which is a huge security risk for the enterprise. Just because you can see the consumer benefit doesn't mean you can use it in the enterprise.

Focusing on the data

The data conversation ends up being important because Salesforce research has found that the average enterprise has over a thousand silos of data across apps. Enterprises need to be intentional about bringing the data together. Teams need to look at and prioritize use cases systematically. O’Sullivan notes:

It's really important that we don't build out five hundred use cases all at once because we could potentially burn through budgets too quickly without seeing a return. Otherwise, we will overwhelm ourselves with some of the complexities. And that's human nature. So, we've got to keep it really simple and connect to some of this stuff. So, we can draw an outcome really quickly.

The Salesforce Data Cloud plays a big role in helping to harmonize data from different sources into a unified view. The company has also recently announced the general availability of its vector database that connects and aligns unstructured data from PDFs, emails, transcripts, and other formats with structured data in the data cloud.

O’Sullivan believes the combination of better training and data infrastructure is planting the seeds for more measured but sustainable innovations in the long run:

I think the future looks like the connection of multiple small different innovations coming together. I think we won't see one big bang thing. Yes, we're going to see many more different language models, algorithms, and models that have been trained in different ways, more complex neural networks, etc. Hopefully, we'll see new business opportunities that will create a positive disruption in existing industries.

I've been saying this for now, probably five years. So, this, this is where I'm probably going to get it wrong, but we're going to see some convergence and deeper overlaps in industries. So, if you take a self-driving car, who does the insurance company insure in the future - the manufacturer of the car, the driver behind the wheel, or the software engineer who wrote the algorithm? Or do they insure all three and split the risk profile slightly differently? I think these things are things that we have to solve for. But I think it will be very interesting to see how we can drive this cross-industry disruption and innovation more broadly.

My take

AI hype makes impressive clickbait for mainstream media by glossing over its practical limitations and exaggerating its risks. Salesforce’s efforts to promote more practical training, conversations and collaboration will go a long way toward the more measured progress required for AI efforts to succeed in the long run.

Loading
A grey colored placeholder image