2864: Gigster And The AI Manufacturing Evolution

[00:00:00] Is the future of manufacturing quietly being reshaped by the silent gears of artificial

[00:00:08] intelligence? Well today we're going to step into the world of AI driven manufacturing. A

[00:00:14] realm where efficiency meets personalisation and where the traditional boundaries of production

[00:00:20] are being redrawn. Because joining me today is Humberto, Principal Solutions Engineer

[00:00:26] at GIGSTER and he's also a seasoned expert with over 15 years of experience in the software

[00:00:32] industry including notable stints at Cloudflare and in Moby. But Humberto stands at the forefront

[00:00:39] of integrating AI into manufacturing, turning the futuristic into the now.

[00:00:46] And our conversation today will not only highlight how AI is revolutionising the manufacturing

[00:00:51] process but also delve into the ethical implications and that symbiotic future of human and machine

[00:00:59] collaboration from predictive analytics to co-bots. We're going to unpack the layers

[00:01:05] of AI's transformative power in manufacturing today.

[00:01:09] Now before I get today's guests on, it's time for me to mention the sponsors of

[00:01:14] Tech Talks Daily and in an era where digital security is non-negotiable, legacy managed

[00:01:21] file transfer tools, they simply don't cut it now. So that's where KiteWorks comes in.

[00:01:26] Revolutionising the MFT landscape with unparalleled security credentials including the much coveted

[00:01:33] FedRAMP moderate authorisation. This isn't just about compliance though, it's about

[00:01:37] offering a secure efficient platform for today's remote workforce. So with KiteWorks

[00:01:43] you can benefit from advanced file sharing, email security and customisable integration

[00:01:49] all within a platform designed to safeguard your most sensitive data. So don't let outdated

[00:01:55] technology compromise your security, step into the future of secure managed file transfer,

[00:02:00] get started today by going to kiteworks.com, that's kiteworks.com where security meets

[00:02:07] sophistication. But now it's time to get today's guests on. So buckle up and hold

[00:02:12] on tight as I beam your ears all the way to San Francisco where today's guest is waiting to join us.

[00:02:20] So a massive all welcome to the show, can you tell everyone listening a little about who you are

[00:02:26] and what you do? Yes of course Neil, so I lead solutions engineering for Geekster. Now Geekster

[00:02:33] is a smart software development platform based around the freelancer network and for them I

[00:02:38] work on solutions engineering which means planning, architecting and orchestrating

[00:02:45] basically the technical backgrounds for our customer projects. I've been working around 15 years in the

[00:02:52] software space including work at in Moabee and Cloudflare and in particular now my focus is on AI

[00:03:00] machine learning data and other advanced solutions. And as you said that you got a lot

[00:03:05] of experience not only at Geekster but also your previous roles as Cloudflare and in Moabee etc.

[00:03:11] But I've got a list for some of this that understands that AI is not something that's

[00:03:17] completely new here that just arrived 18 months ago and you've seen everything evolve.

[00:03:22] How do you see AI reshaping the foundational aspects of things like software development

[00:03:28] and its integration into other industries particularly manufacturing? Now there seems

[00:03:33] to be this mainstream adoption available. Definitely, definitely so now we see ever more

[00:03:41] levels of data more signals both coming from consumers themselves as well as more signals

[00:03:47] that and data available from manufacturing processes and supply chains as well.

[00:03:53] So what this leads us to is more opportunities one for optimization,

[00:03:58] second the ability to build products that are already to a degree AI aware, AI ready,

[00:04:06] able to connect themselves with different systems. And it also allows for different

[00:04:12] ways to get closer to the consumer for companies that are involved in manufacturing to be able to

[00:04:18] adapt and anticipate to consumer needs to a degree.

[00:04:22] And the reason I mentioned manufacturing a few moments ago before you came on the podcast

[00:04:26] I was doing a little research on you and I was reading your article where you discussed the critical

[00:04:32] stages where AI can significantly impact manufacturing in particular. So anyway,

[00:04:38] let's obviously listening not say that article. Can you expand on how companies are leveraging

[00:04:43] AI to navigate the balance between mass production and that increasing demand for

[00:04:47] personalization because it is quite a tricky balance isn't it?

[00:04:51] It definitely is because traditionally mass production requires a lot of planning for capacity,

[00:04:58] planning for production, planning for logistics and so on as to be able to do that at scale

[00:05:05] and also anticipate consumer demand which are always shifting can be complicated. So

[00:05:13] to do this effectively I think it's a combination of making sure that some of the key

[00:05:19] manufacturing steps are already set in place but also allowing enough flexibility to introduce

[00:05:27] customization at certain levels of the production and distribution cycle. And one of the advantages

[00:05:35] that data and AI provides is that there's a lot more leading indicators that start to give you

[00:05:41] an indication of where consumers are going. So with predictive analytics, you have a lot better

[00:05:47] idea of where consumer trends are going, what's going to be popular in the fall even in the summer

[00:05:54] and also to be able to plan ahead for that because it's always better to be able to anticipate

[00:06:00] needs than to have to go out of the way at the last moment to fulfill them.

[00:06:06] And the other thing that's changed in manufacturing that's also tied into this is that

[00:06:11] consumers are obviously consuming a lot more services and even digital goods that inherently

[00:06:20] live in the digital space. But obviously there's also more traditional physical goods and this

[00:06:27] area of overlap where there can be a combination of a bundle of physical and digital as part

[00:06:35] of one offering and obviously there's a manufacturing component there and there's also more of a

[00:06:40] digital component there and that's where data and AI can really help to optimize both.

[00:06:48] And of course the concept of co-bots working alongside humans, complimenting workers not

[00:06:54] replacing them is incredibly exciting to me and it also represents this fascinating evolution in

[00:06:59] manufacturing. So if I was to ask you to look into the future through a virtual crystal ball,

[00:07:05] how do you envision the future workplace with such collaborations like that and

[00:07:10] are there any implications that you think it might have on workforce development and training

[00:07:15] and those kind of areas too? Definitely and I actually think this is one of the most

[00:07:21] interesting areas of this whole trend because there's a really unique opportunity for

[00:07:26] mutual learning in this with a co-bot scenario because both humans and sort of the artificial

[00:07:34] helpers each have their own strengths and weaknesses. And so things like Generative AI for example

[00:07:42] can help accelerate learning, can help prepare a learning curriculum for let's say a human needs

[00:07:49] to learn a new part of a production process and needs to be able to learn it fairly quickly

[00:07:54] and there's no real time to go out and get a human tutor and so on. So in that way the co-bot itself

[00:08:02] to a degree or the let's call it the learning extension of the co-bot can help the human learn

[00:08:08] and at the same time this notion of co-bots they basically will learn from the accumulated

[00:08:17] experience of human workers right? So what is worked let's say in a manufacturing process

[00:08:22] or what is worked in doing pretty much anything within a process and that becomes humans teaching

[00:08:31] these co-bots. And so that becomes a really interesting collaboration where you know co-bots can

[00:08:38] learn from historical best practices and humans with their extra context are able to obviously

[00:08:44] add their own particular skill sets but you have a mutually beneficial relationship where both can

[00:08:53] contribute. And Generative AI has been highlighted by so many experts as a game changer for

[00:09:01] manufacturing enabling rapid prototyping and ideation and so many great things but of course

[00:09:07] when we start talking about Gen AI the next thing I've got to mention is given that

[00:09:11] legal challenges around things like intellectual property combined with Gen AI as seen with

[00:09:18] so many others out there I'm curious what are your thoughts on but navigating these emerging

[00:09:24] legal and ethical considerations especially when we're talking about responsible AI it's a huge

[00:09:30] topic right now but I'm just curious what your thoughts are on this. Yeah this is an interesting

[00:09:36] and challenging question because it becomes a little bit of a cat and mouse game so technology

[00:09:43] always tends to outpace regulation and laws and so on. And so right now we're at an era where

[00:09:52] from a technical standpoint there's the possibility of generating all of these digital

[00:09:58] goods and then the digital goods that can become the molds for physical goods that are

[00:10:03] based around things that really have IP protection around them and that perhaps

[00:10:10] should not be produced and how do we protect that. But I think the advantage there is also that one

[00:10:18] some of this technology that's used to replicate can be turned around and used to detect as well.

[00:10:25] So using artificial intelligence as well and as a computer vision or other forms of analysis

[00:10:33] is actually can become possible to quickly detect let's say cases of intellectual property

[00:10:40] infringement in the wild be it on an e-commerce website or obviously this is already being

[00:10:47] done with things like streaming video and audio that this is already in place but

[00:10:52] but on a more expansive way it can be done. And where it also gets might get interesting is

[00:10:58] that it might allow for more efficient than expedient licensing for example let's say a case

[00:11:05] where let's say you want to build something or to put out something and let's say an

[00:11:11] manufacturing standpoint that does need to be licensed and let's say you scan what you're

[00:11:18] planning and it says actually this design is partially owned by X company. It might allow

[00:11:24] in the future for manufacturers to say okay well let me get a quick sub-license to actually produce

[00:11:30] this in the right way or things of that nature. So I think we're going to get to a

[00:11:36] world where through new forms of attribution supported by General AI and other technologies

[00:11:43] it can become efficient and more creative in a way that also respects the rights

[00:11:50] of rights holders from an AP perspective. Yeah I completely agree with you especially

[00:11:55] around the efficiency point now and I think with the AI's continuously growing influence on the

[00:12:00] manufacturing sector there's also something we don't talk about enough and that is the

[00:12:05] potential for a significant environmental impact through more efficient production processes and

[00:12:11] reduced waste. Is this something that's important to you at Giggs2 and if it is what's your

[00:12:17] approach to sustainability in those AI driven projects? That's a great question and it definitely

[00:12:24] is important to us and we look at it from several angles. One is that inherently

[00:12:33] processes that are streamlined are inherently more efficient and that leads to an energy

[00:12:39] saving wherever those processes touch energy consumption. If you think about

[00:12:46] let's say in a manufacturing process to the degree that you can reduce let's say unexpected

[00:12:52] spikes in demand that can trigger let's say you need extra fast shipping by plane where

[00:13:00] if you had planned ahead of time you could have shipped by ship or rail and that is drastically

[00:13:06] potentially more efficient then that's an environmental plus right there. The other side

[00:13:13] of things is from the AI perspective itself. So there's been technologies recently that have been

[00:13:20] inherently they consume a lot of energy. Obviously blockchain and other technologies have been

[00:13:25] particularly highlighted for energy consumption and as far as AI in particular the training stages

[00:13:32] of large models are very consumption intensive and so to the degree that part of our work involves

[00:13:45] looking at customer needs and saying well what do you actually need at the proper scale and we're

[00:13:51] not going to be doing large-scale AI for AI's sake and building gigantic models just to build

[00:13:57] them then that is also that also has a positive AI the positive impact from this kind of environmental

[00:14:07] standpoint of saying well we're going to reuse a model that already exists and therefore

[00:14:12] you're going to use it for inference which itself is a very low impact comparatively speaking

[00:14:19] mode of energy use and so on. So the more we actually optimize AI the use of software,

[00:14:26] the use of AI itself that also leads to a positive impact there.

[00:14:32] And I also wanted to highlight that the delivery phase of manufacturing is also under going somewhat

[00:14:37] of a revolution right now especially with AI's helping things like demand forecasting and supply

[00:14:43] chain optimization but again from your perspective what are the next frontiers for AI and ensuring

[00:14:49] that products not only reach consumers more efficiently but also align with their changing

[00:14:55] needs because again a big talking point right now isn't it? Definitely and I think there's two

[00:15:01] sides to this. One is that as one of the things I was mentioning earlier is there's ever more signals

[00:15:08] and touch points with customers right there's customers at the point of sale, there's customers

[00:15:13] on the web, there's communities that customers interact with online that influence demand

[00:15:19] requirements and so on. So the ability to ingest this information within a data system that involves

[00:15:27] AI and is able to take this information and provide feedback to product and to manufacturing

[00:15:35] that helps align with consumers changing needs because you can see their potential use of a

[00:15:41] product or their attitude towards a product or their needs change almost in real time.

[00:15:48] And as we were talking about earlier this notion of testing and trialing in batches

[00:15:54] and allowing for customization also allows this adaptation to changing needs.

[00:16:02] And finally, well obviously already in April if we dare to look beyond 2024 I know it's early

[00:16:09] still yeah but are there any other emerging technologies or indeed trends that you think

[00:16:14] will further revolutionize that manufacturing industry and is there anything you can share around

[00:16:20] or any teasers you can share around how PIXTER is preparing to be at the forefront of some of these

[00:16:25] changes too? Right, right definitely so there's a lot of interesting changes that seem to be on

[00:16:33] the horizon and quickly coming to the field. So the notion of pervasive AI

[00:16:41] is really coming into the fold. This notion that AI is not only happening

[00:16:49] centrally in a particular, you know, in the cloud or someplace else but AI also happening

[00:16:57] at the edge within our own devices, be it within our phones or within devices that exist physically

[00:17:04] and so AI being more pervasive all around is something that's definitely coming.

[00:17:11] And along with this are the effects of data flowing seamlessly across supply chains.

[00:17:18] So this notion of data silos continuing to splinter and really allowing for data to flow

[00:17:26] to where it needs to to be able to act intelligently at whatever stage of let's say a production

[00:17:33] cycle or even at the level of consumer use is something that's going to be important.

[00:17:41] And overall one of the things that's also key is that everybody is going to be made smarter by AI.

[00:17:48] This includes consumers and producers and so we're all going to be living in this new world

[00:17:54] where we're all being helped by AI to be smarter and I think that'll change a lot

[00:18:00] of things in manufacturing and beyond. Well you've shared so many golden insights today

[00:18:07] and before I let you go though I'm going to ask you really one final gift with everybody

[00:18:11] listening around the world today and that is a book to add to our Amazon wishlist.

[00:18:16] So I always ask my guests to leave something behind that other people listening can check

[00:18:20] out maybe it'll inspire those but maybe it can inspire them but what book would you like

[00:18:24] to add to that wishlist and why? Well thank you and I appreciate that question and what I would

[00:18:31] like to leave our audience with is a classic that I've very much liked for many years it's

[00:18:39] Carl Sagan's The Demon Haunted World and it's a book that really is champions the scientific

[00:18:46] method and critical thinking as a counterpoint to shallow thinking and superstition and so on.

[00:18:55] And this kind of a skeptical scientific view as really a way to understand the world better

[00:19:03] and I think that can be a very broad applications and it's one of my favorites.

[00:19:09] Absolutely love that. Well thank you so much for joining me on the podcast today and for

[00:19:14] anyone listening just wants to find out more information about yourself maybe contact your

[00:19:18] team or we'll just learn more about Gixxster and how you're helping businesses and manufacturers

[00:19:23] around the world what's the best starting point for everything? That's a good question definitely

[00:19:30] our website at Gixxer.com so well I'll add links to that I might even add a link to your

[00:19:36] LinkedIn as well just in case anybody has got any questions they'd like to ask you personally and

[00:19:42] just listening to your conversation today reminded me of I think it was who was it now

[00:19:48] looking back I think it was Henry Ford and he said any customer can have a car painted any color that he

[00:19:55] wants as long as it's black because back then they could get Ford could get away with that stuff this

[00:20:00] lack of personalization because manufacturing processes gave them such a competitive advantage

[00:20:06] and made that price of the car so cheap but 100 years later manufacturers now need to be doing

[00:20:12] both don't they they need a more customizable personalized products and they need streamlined

[00:20:18] efficient processes to make it and as a result manufacturers are increasingly turning to AI and

[00:20:24] listening to you break that down today those three manufacturing stages where AI can help the most in

[00:20:30] things like consumer demand for customization ideation and testing and production I think

[00:20:35] you brought it all together beautifully today so just a big thank you for sitting down and sharing

[00:20:40] your story today well thank you for so much for having me on after my conversation with

[00:20:45] Humberto today I think it's clear that the journey of AI in manufacturing is not just about

[00:20:51] automation it's so much more than that it's actually about augmenting human potential and

[00:20:57] pushing the limits of what's possible and his insights shared today reveal a manufacturing

[00:21:03] future that is not only more efficient and sustainable but one that also aligns closely with

[00:21:08] the nuances of consumer demand so everybody wins and AI's role in manufacturing that's also evolving

[00:21:17] heralding this new era where machines learn from humans and humans learn from machines creating this

[00:21:24] tapestry of innovation that is bound almost to redefine the landscape of manufacturing

[00:21:30] but after this conversation I find myself pondering the endless possibilities of

[00:21:35] what AI can bring to the table and also wonder what new horizons will AI open next in our relentless

[00:21:42] pursuit of progress and personalization in manufacturing so I invite you to share your

[00:21:47] thoughts and join this conversation and how we can navigate this exciting frontier together

[00:21:53] so listeners what's your take on the future of AI and manufacturing how do you see AI shaping

[00:21:59] our world in the years to come please share your thoughts let's keep this dialogue going by

[00:22:03] simply emailing me techblogwriteroutlook.com twitter linked in instagram just at neil cqs

[00:22:10] and if you enjoyed today's episode join me again tomorrow we've got another topic

[00:22:14] that I also invite you to join me in discussing other than that just thank you for listening

[00:22:20] as always really appreciate your time and until next time don't be a stranger

[00:22:33] you