What Was Everyone Talking About at Nasuni CloudBound 2024?

Whoever said, “What happens at CloudBound, stays at CloudBound,” clearly has never attended a Nasuni-hosted event. In fact, we love when people share their learnings, especially if it can help other companies achieve the many benefits of the cloud.

Over 800 registrants across the globe – including representatives from notable brands like Dow and Mattel – were invited to engage in a packed agenda of talks fireside chats, and testimonials from a selection of 11 renowned speakers.

And, frankly, that’s just too much goodness to keep quiet about.

So, what was the word on everybody’s lips? Well, no prizes for guessing. It was: ‘AI’. From superpowers and rocket fuel, to risks and revolutions, the many faces of AI strategy were top of mind for everyone at this year’s event.

We were keen to approach the subject from some novel angles, which is why we put together an all-star lineup of respected voices in the field – among them the first woman to lead MIT’s AI Lab, IT leaders from around the world, and an ethical hacker. At Nasuni, we’re not about adding needless volume to the noise around AI. Instead, we want to cut right through it. With that in mind, here are my top highlights from the day.

“We don’t work on today. We don’t work on tomorrow. We work on the day after tomorrow.”

I was lucky to get the chance to sit down with Professor Daniela Rus, director of MIT’s Computer Science and Artificial Intelligence Lab, to discuss the risks and rewards of the AI tsunami we’re all riding.

She recognizes how fantastic it is that generative AI has democratized AI – putting it in our minds and in our pockets. Daniela also reinforced the importance of the public understanding that GenAI is not the only flavor of artificial intelligence available. She believes that even greater potential lies in combining different kinds of AI to empower humans to make better decisions.

Daniela also laid out a major challenge facing developers using AI to design prototypes: AI doesn’t understand the laws of the physical world. To get around this, her team has been experimenting with connecting a Stable Diffusion-like algorithm with a simulation engine. They pass the designs produced through the simulator to check they meet the physical specifications. By iterating through this, they’ve been able to build prototypes that function as expected in the real world. As a result, they’ve reduced the timeline from concept to working machine from months to hours.

“I ask two questions: ‘What’s difficult?’ and ‘What’s cool?’”

At Nasuni, we collaborate with all our customers to help them get maximum value from the Nasuni File Data Platform. So it’s always great to hear how our solutions are being used in the wild. Jim Liddle, Nasuni’s Chief Innovation Officer of Data Intelligence and AI, sat down with Charles Douglass, Director of Information Technology at McKim & Creed, to learn how they are putting AI into action.

McKim & Creed currently have three different AI use-cases. They’ve developed an HR chatbot using Microsoft Copilot to answer employee questions. Alongside that, they’re using AI services to assist with proposal writing by extracting insights from historical proposal data. Their third use case is already widespread among our customers: automated meeting notes with recommended action items.

But they’re not resting on their laurels. Their future plans include leveraging their 150+ terabytes of Nasuni-managed data to develop AI business tools including project summarization, project timeline generation, and financial analysis. They’ve already begun exploring ways to structure this data for effective AI use and maximum automation.

Charles’ advice for companies that are still unsure about AI was clear: just start. Begin experimenting with AI even without a clear end goal. Focus on solving user pain points, get to know the fundamentals of how it works, and think creatively about potential applications in your own industry.

“For a long-term AI strategy, you really need an AI data strategy.”

Our very own Jim Liddle took to the virtual stage to give an intensely practical run down of how to lay the groundwork for AI success.

AI needs a single source of truth and knowledge. When unstructured data is stuck in silos, it becomes hard to access and analyze. Furthermore, unstructured data runs the risk of becoming ‘dark data’, after which it becomes hard to locate and even harder to access. This combination of dark data split across silos flies in the face of AI readiness. To avoid this scenario, Nasuni has developed a five-step ‘Fit for AI’ framework, which can be viewed in more depth here.

It always starts with an assessment of the unstructured data silos outside of Nasuni, mapping out what is where, and giving thought to potential AI use cases. Next comes consolidation. The silos are brought over to Nasuni, providing a unified view of that unstructured data. This transparency improves the discoverability of the data sets, allows for more efficient processing, and boosts scalability.

At this stage, we recommend having another security assessment. Make sure your user permissions are correct, check that access control lists are up to date, and get a solid disaster recovery plan in place. Finally, comes the most important piece of the puzzle: curating the data. This can include anonymization and masking of personally identifiable information (PII) data, setting data quality goals, and planning how the data will be used and which analysis it will drive.

For me, the overarching takeaway from Jim’s presentation was that a solid data strategy is the key to a cohesive, long-term AI strategy.

“Your adversary is a good teacher because bad guys show us where technology is going.”

Ethical hacker, Keren Elezari, closed out the event with some surprising insights about what we (the good guys) can learn about data security from the murky underworld of cybercrime.

She revealed that ransomware made over a billion dollars in revenue last year, making it the most successful form of cybercrime on the planet. And that’s just the payment of the ransom itself – not including the damages, downtime, or how much it costs a company to recover. Incredibly, the fine print of some extortion notes also includes job ads. They’re recruiting from amongst their victims!

Keren pointed out that, since generative AI has been on the scene, cybercriminals have been using it as rocket fuel for their innovations, taking advantage of malicious GenAI and LLMs. There are three main ways that GenAI is helping hackers do what they do best, only faster: 1) translating and localizing text for phishing emails, 2) harvesting credentials, and 3) cracking passwords, and directly scanning and exploiting network connectivity technologies and VPNs to infiltrate target systems.

She left us with an uncomfortable question: should we all adopt a “hacker mindset” to stay ahead of the incredible innovation and adaptability of today’s cyber criminals?

And with that, the curtain came down on CloudBound 2024.

September has seen a whole series of online and in-person Nasuni CloudBound events all over the world – from California, to Boston, to Chicago, to London. It’s been a whirlwind of valuable information and discussion, but the fun doesn’t stop there! We have two more events in October in New York and Philadelphia, so it’s not too late to experience how hybrid cloud can work for your business AI strategy. If you missed it, or want to relive it from the comfort of your home, you can catch up on all the talks here.