The AI Factory at Dell Technologies World 2024

3:54 · Watch on YouTube ↗

Transcript 673 words · about 4 min to read

Auto-generated captions from YouTube, not hand-corrected, so names and technical terms may be imperfect. The video is authoritative.

Dell Technologies World 2024 was definitely the AI Edition last year it was the Apex Edition this year it was definitely all AI jinten Wang and Michael Dell standing on stage talking about running huge numbers of broadb gpus in a single Rack in order to deliver business value from large language model Ai and we saw Dell talking about the Dell AI Factory and the AI Factory is a really important thing it's the idea of the industrializing a of large language model AI that we're moving from a

stage where it's a science project where you have to know all of these things about embeddings and Vector databases and chaining and specialization and fine tuning and these are all pieces of knowledge that aren't actually business differentiating for most organizations and so we need to move from where large language models are science projects in order to deploy towards the point where it's an industrial process where we can simply deploy a large language model tool we can bring our own corporate information our corporate data that we

wish to retain and to protect as well as our business requirements our business problems that require solving we just bring these things these are the things we need to know about because they're what differentiates our business we don't need to we don't want to learn about all of the insides of building a large language model AI to solve these problems we just want these problems solved the moment it's definitely a science project and we saw some awesome science projects at Dell technology World in particular was

impressed with the Northwestern University Medical who are using a large language model AI to start this the to build the scaffolding of a radiology report so a xray is taken of a patient the large language model can identify features that are not about the disease that the the patient has so orientation uh of the image maybe uh things like um drains and laws you know medical pieces that are that are there on the person or EKG leads these things can be identified by the LM so

that the radiologist uh doesn't have to spend all of their time doing these routine processes and they can focus then on what's the actual disease what's the thing that I want to see in this image and this is my point is we're bringing our business data in this case an x-ray image and our business problem this this is that we we need to analyze these things and we need a specialized human to do this but simplifying that process for the human getting real business productivity improvements

the radiologist now can actually scan more images or spend more time looking at the disease element of the image than they did previously and it's this ability to industrialize the use of large language models which is what we need to see it's not quite there the Dell AI Factory is more of a concept and an aspiration than it is a shipping product I didn't see any signs that they were actual factories building AI Solutions in any of the AI the Dell products but we will see

this happen over time we're going to see vendors building AI into their product rather than the AI being the product and in particular we've seen this already with things like image recognition as a part of loss prevention or security safety uh products that are monitoring video cameras instead of having humans monitor them uh we'll see the same thing progressively happen with large language mod a you'll see lots more coverage of large language models and Ai and getting business value out of these large language models and

AI here on the CTO advisor you'll also see a lot of coverage through the wider futurum group so check out the futurum group.com as well follow us here on the CTO advisor and we will see you on the next CTO dose