Insights Into Things Podcasts

00:00:02:16 – 00:00:07:06
Narrator
Insightful podcast.

00:00:07:08 – 00:00:14:23
Narrator
By informative hosts for. All.

00:00:14:25 – 00:00:23:29
Narrator
The insights into things through a podcast network.

00:00:24:01 – 00:00:51:15
Narrator
Welcome to insights into technology, a podcast exploring the latest in computers, networking, home automation, mobile computing and all things technology related. Our hosts will take a deeper dive into the latest and greatest in tech trends, and give you the information you need to enable your tech centric world.

00:00:51:18 – 00:01:21:18
Joseph
This is insights and technology. Episode 23 bugs, Breaches and Bot Takeovers. I’m your host, Joseph Whalen. And before we get into your Tech News of the week, I did have an apology and a notification to throw out there. The apology is because our regular release of, podcast was delayed this weekend. We had a bit of a, incident, at my house.

00:01:21:20 – 00:01:50:15
Joseph
As you guys know, my studio’s in my house. We had a situation where our sump pump had gone. We’d gotten some rain. We had gotten some water. We had some, mold issues coming, as a result of that. So we’ve got a significant amount of construction going on. And last weekend, they were in to finish up the work, and we had to, move some stuff down in the room where the mold was.

00:01:50:18 – 00:02:13:06
Joseph
And as a result, I had to take my entire network offline. So I was kind of working outside my comfort zone there and didn’t have a network to work with. I got things back up and running to get the podcast out late on Monday, so I wanted to apologize for that. But the work isn’t done yet. So we’re probably going to have to do another shutdown in the near future.

00:02:13:06 – 00:02:35:17
Joseph
Hopefully I don’t have to shut everything down. But in the coming weeks, as they, contractor come in to finish up that work, there’s a chance that we may miss a podcast deadline like we usually do. So I just wanted to put that out there so people weren’t wondering what happened to us. Anyway, on to our Tech News of the week.

00:02:35:20 – 00:03:22:26
Joseph
Our first story comes to us from TechCrunch. Sarah Perez wrote this article. Autonomy at the expense of privacy. Signal President Meredith Whittaker has raised alarms about the security and privacy risks associated with a genetic AI, or autonomous systems capable of performing tasks without human intervention. Speaking at South by Southwest 2025, she cautioned that AI requires extensive access to personal data, including browsing history, financial information and calendars, potentially operating with root level permissions.

00:03:22:28 – 00:04:09:12
Joseph
This data is often processed offsite, transmitted to cloud servers, thereby heightening security and privacy vulnerabilities. The rise of a genetic AI presents significant challenges for commercial and enterprise computing, particularly concerning data security and privacy. Energetic AI systems necessitate broad access to sensitive data, potentially leading to unauthorized data exposure. Enterprises must evaluate the extent of permissions granted to AI applications to safeguard confidential information.

00:04:09:15 – 00:04:48:11
Joseph
The reliance on cloud servers for data processing by a genetic AI increases the risk of data breaches. Organizations should consider implementing robust encryption protocols and exploring on device processing alternatives. To mitigate these risks. Deploying a genetic AI may complicate adherence to data protection regulations like GDPR and CcpA. Companies must ensure that AI operations align with legal requirements to avoid penalties and maintain consumer trust.

00:04:48:13 – 00:05:22:26
Joseph
This is kind of a big one here in going through the process to become Cmmc certified. There’s a lot of rules with that and many other regulations involving where your data can be, how it has to be handled and who can can actually handle it. And in the situation here where you might not necessarily know what cloud service or whose cloud service the AI is using, it makes it very difficult to maintain compliance moving forward.

00:05:22:28 – 00:05:52:18
Joseph
Clear communication with users regarding data usage by a genetic AI is essential. Assuming you have the information, I think a lot of times you might not even have that information to be, transparent with. Implementing transparent consent mechanisms can help maintain trust and comply with ethical standards. The integration of a genetic AI necessitates a reevaluation of existing security and infrastructures.

00:05:52:21 – 00:06:21:03
Joseph
Organizations should invest in advanced security measures to address the unique challenges posed by autonomous AI systems. It’s important to note that I. It’s been around for a while. In fact, today we’re going to start our deep dive, into the into AI and its history. So it’s been around for a while, but it’s not really taken off for, the general population until just recently.

00:06:21:06 – 00:06:49:12
Joseph
So a lot of people don’t understand what the security implications are or how to fit AI into existing security infrastructures. And that’s going to be, a concern moving forward with organizations where you’re going to have to maintain or continue to maintain compliance with regulations. But as with any new technology, you need to understand the technology before you start deploying it.

00:06:49:15 – 00:07:14:03
Joseph
A lot of companies want to just jump on the bandwagon because AI is this great and wonderful thing. And as it’s being portrayed in the media and they don’t take the time to understand what the risks are and how to best handle it, and to to maintain compliance. So this one should be interesting to see how it pans out in the long run.

00:07:14:05 – 00:08:01:26
Joseph
Next up, Lance Whitney tells us from a ZTE net when X hits the fan on March 10th, 2025 X, formerly known as Twitter, experienced widespread outages affecting tens of thousands of users globally. Elon Musk, owner of X, attributed these disruptions to, quote, a massive cyber attack suggesting involvement from a coordinated group or nation possibly originating from Ukraine. Concurrently, a pro-Palestinian hacker group called the Dark Storm Team claimed responsibility for the attack, stating it was in response to the ongoing conflict in Gaza.

00:08:01:28 – 00:08:36:00
Joseph
Users reported issues such as slow loading times and error messages, with the most significant impact observed around 10 a.m. Eastern Time. The cyber attack on X underscores significant vulnerabilities in digital infrastructure, highlighting potential risks for businesses relying on social media platforms for communication and marketing. Now, you know, there’s there’s certainly platform vulnerability issues that have to be taken into account here.

00:08:36:02 – 00:09:02:21
Joseph
Information was not very forthcoming from X on this, but the only attribution that Musk had given to this being a Ukraine issue was they traced it to Ukrainian IP. So a five year old would make a conclusion like that, spoofing IP addresses or not even spoofing them, just using VPN ends to appear to be coming from somewhere else is trivial at this point in time.

00:09:02:23 – 00:09:36:24
Joseph
So if the solid evidence they have is just the IPS were Ukrainian, I have to seriously question what the origination point of the attack was. The attack does reveal potential weaknesses in widely used platforms, emphasizing the need for robust cybersecurity. Businesses should assess their reliance on third party platforms and develop contingency plans for potential disruptions. Fortunately, X isn’t the most widely used social media platform right now, certainly not by businesses.

00:09:36:27 – 00:10:12:00
Joseph
So I think the business impact of the commercial impact or enterprise impact was minimal in this case. Here. Talking about the attribution challenges, Musk attributed the attack to Ukrainian sources, which contrast with the Dark storm team’s claim, illustrating the complexities in accurately identifying perpetrators. It was really premature to come out and claim anything. When you don’t, you’re you have an ongoing investigation and you don’t have any solid evidence of anything.

00:10:12:02 – 00:10:41:17
Joseph
But the ambiguity complicates response strategies and underscores the importance of comprehensive threat intelligence, which clearly we we got nothing from this one. So you take it at face value. And I don’t think that’s worth all that much in this case. The outages on platforms like X can hinder real time communication, affecting customer engagement and service delivery. Enterprises should diversify communication channels to mitigate such risks.

00:10:41:17 – 00:11:07:00
Joseph
And and this is a good thing no matter what because not everybody is on a single platform. So if you’re going to use social media as a preferred method of communicating with customers or clients or vendors or whoever, you probably want to do it across multiple popular platforms so that the next time someone decides that they want to take X down, it doesn’t affect your ability.

00:11:07:02 – 00:11:40:26
Joseph
The prolonged platform outages in this case can lead to reputational damage for businesses perceived as unresponsive, so if you’re primarily communicating through a single entity like this and it goes down and you go dark with any announcements or information that you have to just, disseminate, then their failure starts to look bad on, on your part. And that’s probably something you definitely want to avoid as a, as a commercial company.

00:11:40:28 – 00:12:07:11
Joseph
Next up, Cyber Scoops Tim Starks tells us, from Coast Guard to Cyber Guard. In a significant move for national cyber security, President Trump has nominated Sean Plank to lead the Cybersecurity and Infrastructure Security Agency, or CSA plank. He brings a wealth of experience to the role. He served as principal deputy assistant secretary, not just roles. Right off the tongue.

00:12:07:11 – 00:12:42:28
Joseph
That title does it for the Department of Energy’s Office of Cybersecurity, Energy Security and Emergency Response. Between 2019 and 2020. Prior to that, he was director of cyber policy at the National Security Council. Starting in 2018. I don’t know if that’s a wealth of experience. That’s literally two years of experience, which really makes me wonder his qualifications like he is a formal former US Coast Guard officer, having held positions such as deputy division chief.

00:12:43:00 – 00:13:23:28
Joseph
His service record includes time in Afghanistan and work for U.S. Cyber Command. If confirmed by the Senate plank, he will inherit an agency facing new challenges, many of which we’ve discussed on this podcast. Actually. Under the current administration, Cisa has experienced personnel reductions and shifts in its responsibilities, raising concerns about its capacity to effectively safeguard critical infrastructure. Like his nomination carries significant implications for commercial and enterprise computing, the first being leadership stability.

00:13:24:00 – 00:13:48:14
Joseph
His extensive background in cybersecurity and, you know, the article was very generous in saying he has extensive background. I don’t think two years, 2 to 3 years is an extensive background in cybersecurity. I’ve been working in it myself for 30 plus years now, and there’s every day something that I realize I don’t know. So three years, probably in three years as being generous.

00:13:48:15 – 00:14:17:11
Joseph
He started in 2018 and he ran through 2020. So a little over two years, I would say saying that that’s extensive, experience is really not a good idea, although compared to anybody else who’s been put into cabinet positions, that probably has extensive experience, changes in CSIs focus and resources may affect public private partnerships essential for protecting critical infrastructure.

00:14:17:11 – 00:14:43:11
Joseph
We already talked about that staffing reductions could hinder Cece’s ability to respond to cyber threats, potentially increasing risks for businesses. Although they’ve the administration has come out and basically said Cece’s not going to be working with the commercial or private sector moving forward. They’re just going to protect government entities. And in an extreme case, only federal government entity.

00:14:43:11 – 00:15:17:24
Joseph
So state and local kind of have to fend for themselves. So those staffing reductions might not be as significant since the role has been curtailed significantly. Trump, like his background in government, cybersecurity, military, cyber operations and private public collaboration, aligns with Cece’s mission to protect critical infrastructure and national security. Again, that’s being very generous. His experience in policy development and cyber defense, cyber defense strategy positions.

00:15:17:24 – 00:15:47:17
Joseph
I’m sorry, positions him to address evolving threats for his effectiveness. That was. Effectiveness will depend on navigating Cece’s resource and mission challenges. Again, I don’t think we’re talking about a massive amount of experience here, so I don’t know what his policy development experience has been. There’s been very little information released on what he he’s done in his limited 2 to 3 year cybersecurity career.

00:15:47:20 – 00:16:22:27
Joseph
I certainly wouldn’t take someone with that little of experience and put them in a security, a senior security role in my company. I can tell you that right now. Recent changes within CSA, including leadership shifts and evolving cybersecurity policies, signal a stronger federal focus on public private collaboration and regulatory oversight. Enterprises should prepare for increased compliance requirements, enhanced threat intelligence sharing, and a greater emphasis on proactive cyber resilience strategies.

00:16:23:00 – 00:16:51:24
Joseph
Now, it’s interesting that that we we go down that route with that logic. You have to prepare for all of that because Cece’s not going to do it anymore. So if you’re not and I question the increased compliance that I think that we’re still kind of up in the air about that, but enhanced cyber threat intelligence sharing is absolutely a must because you’re not going to be getting it from the government.

00:16:51:27 – 00:17:18:10
Joseph
So hopefully there’s going to be some kind of groundswell of support for better communication to make up for this gap that they’re creating. I don’t know, we’ll see. I mean, I think the heads of these agencies don’t necessarily, and it’s been proven in the past, don’t necessarily need to be experts in what those agencies do. They’re figurehead positions.

00:17:18:15 – 00:17:48:02
Joseph
They’re basically power positions to, I don’t know, position allies in the government. It’s a political move more than anything. These aren’t the people that are actually doing the work. So that’s that’s kind of the only positive takeaway I have from that. So we’ll see how things go. We’re going to take a quick break. But before we do, I want to invite you to, give us your feedback.

00:17:48:04 – 00:18:13:12
Joseph
Let us know how we’re doing. Let us know your opinion on these on these articles that we’re talking about here. Tell me I’m wrong. Tell me I’m right. Let’s have an open dialog about it. I’d love to have, a discussion with someone about this. You can email us at. Comments and insights into things.com. You can call in and leave a voicemail for us at (856) 403-8788.

00:18:13:15 – 00:18:38:09
Joseph
That is (856) 403-8788. Or as always, you can get links to all that and more on our official website at W dot insights into things.com. We’ll be right back.

00:18:38:12 – 00:18:56:12
Narrator
Insights into entertainment a podcast series taking a deeper look into entertainment and media. Our husband and wife team of pop culture fanatics are exploring all things, from music and movies to television and fandom.

00:18:56:14 – 00:19:40:21
Narrator
We’ll look at the interesting and obscure entertainment news of the week. We’ll talk about theme park and pop culture news. We’ll give you the latest and greatest on pop culture conventions. We’ll give you a deep dive into Disney, Star Wars and much more. Check out our video episodes at youtube.com. Backslash insights into things. Our audio episodes at Podcast Insights into entertainment.com, or check us out on the web at Insights into things.com.

00:19:40:24 – 00:20:20:22
Joseph
Welcome back to insights into Technology. Our next article comes to us from Ann Thompson at The Register. When developers go rogue, a federal jury in Cleveland has convicted Davis Lou, a 55 year old senior software developer from Houston, Texas, of sabotaging his former employers, IT systems. This is a this is a real nightmare scenario here. Employed by Eaton Corporation from November 2007 to October 2019, Lou faced the motion during a corporate restructure in his final year.

00:20:20:24 – 00:20:58:13
Joseph
In retaliation, he introduced custom malware designed to crash production systems by consuming excessive resources. Additionally, he implemented a quote, kill switch that locked out thousands of employees globally. Upon his termination on September 9th, 2019, causing significant operational disruptions and financial losses. The incident underscores the critical importance of robust internal security measures to protect against potential insider threats within the organizations.

00:20:58:15 – 00:21:27:10
Joseph
Now, this is insider threats is definitely something that is near and dear to me because it is something that the federal government highly emphasizes the dangers of when it comes to cmmc. In fact, there’s been numerous training sessions that I’ve had to go through, and our our users have had to go through organizations. Most recognize that employees with extensive system access can pose significant risks.

00:21:27:10 – 00:22:04:24
Joseph
If disgruntled, and implementing regular training and fostering a positive work environment can mitigate such threats. Regular audit audits of user privileges and continuous monitoring can help detect unauthorized activities early, preventing potential allergies, providing critical tasks, or dividing critical tasks, not providing dividing them among multiple employees ensures that no single individual has unchecked control, reducing the risk of malicious actions.

00:22:04:26 – 00:22:57:05
Joseph
I have to take a second to kind of delve into that one a little bit. Working in it, it is in a very unique position where invariably we have the keys to the kingdom. We have to have access to everything and the trust that we rely on in order to do our jobs is vital. And in situations like this, when you have an individual who’s in that technology organization, who violates that trust because they’re disgruntled or or for whatever reason, seriously erodes the trust that all companies tend to have in their I.T organizations at that point in time, because they see situations like this and they immediately have a fear factor of it happening to

00:22:57:05 – 00:23:27:19
Joseph
them. And when that happens, they try to start pulling back on access, they start pulling back on permissions, and it makes our job very difficult to do. Now there is very simple, steps that you can take that we discussed in our frameworks, framework. Security framework, deep dives such as least privileged, access, separation of duties and so forth that will help to mitigate this type of thing.

00:23:27:21 – 00:23:55:28
Joseph
You don’t want anybody who has root level or admin level access to be using that kind of power on a day to day basis. So having specialized accounts that can be easily audited and controlled is the probably the best way to do something like that. Developing and rehearsing comprehensive incident response plans enables organizations to react swiftly to internal threats, minimizing damage and downtime.

00:23:56:00 – 00:24:22:24
Joseph
But in. When we talk about insider threats a lot of times and the gathering kind of falls into this, rut themselves with the training they do, they want to go for the people that are stealing government secrets and sending it to the Russians or whoever the bad guy of the day is. A lot of times, your insider threats are much less, obvious than that and much less cliche than that.

00:24:22:26 – 00:24:48:28
Joseph
Most internal or insider threats are in are unintentional. So people that fall for phishing scams, there’s an insider threat. You know, if somebody clicks on someone, tricks your employee to click on something that gives a nefarious actor access to your network, well, that’s an insider threat. It’s not, you know, and some of the things that the training gives you.

00:24:48:28 – 00:25:10:17
Joseph
Oh, well, did Bob, why did Bob buy a new car? Why is he going on vacation? Why is he doing this? Why is he doing that? Why is he, talking to a, you know, a foreign looking person? And why is he taking foreign trips? Insider threat awareness usually is not to that extent in most cases. I’m sure that does happen.

00:25:10:19 – 00:25:33:09
Joseph
But most of your insider threats are things like this, where you have an employee that has all this access and you make them angry, you make them disgruntled, and they feel that in order to get their due, they’re going to get back at you. Now, I don’t say that. That’s justified. Certainly I, I was never, of the mindset of burning bridges.

00:25:33:09 – 00:26:02:20
Joseph
If I leave companies, regardless of what terms I leave the companies on or why I leave them, there’s just no point in. This guy’s facing a potential ten year prison sentence. Now, because he really did something stupid and got nothing out of it. So insider threat awareness is definitely a significant issue to be aware of, but understand that it’s not the typical spy versus spy insider awareness stuff that you usually have to worry about.

00:26:02:22 – 00:26:25:29
Joseph
It’s usually more benign, harder to find. You know, it was nobody knew this guy put these back doors in. So having regular audits, having someone look over everyone’s shoulder, I’m director of it where I’m at, and I’m not above reproach. I invite anybody to come in and make sure that I’m not doing anything wrong, because that’s how that trust is built.

00:26:26:02 – 00:27:03:04
Joseph
You know, questioning my actions isn’t questioning your trust. It’s being smart. And it’s important for people to understand that. Next up, BleepingComputer Serge Catlin again, he’s a popular one that we go to all the time. Chrome Catchers and Android Avengers. In 2024, Google awarded nearly $12 million to 660 security researchers for its Vulnerability Reward program, or VPI.

00:27:03:07 – 00:27:40:22
Joseph
This initiative encourages the discovery and responsible reporting responsible. You have to emphasize that the responsible reporting of security vulnerabilities across Google’s products and services, notably the highest single reward, exceeded $110,000. That’s good work if you can get it. Google’s substantial investment in bug bounties underscores a proactive approach to security, setting a benchmark for other organizations to prioritize vulnerability management by offering significant rewards.

00:27:40:24 – 00:28:21:07
Joseph
Google motivates the global community to identify and report potential threats, leading to more secure software and services. The VP’s expansion to include various platforms such as Chrome, Android, and cloud services ensures that a wide array of products benefit from rigorous security assessments. Google’s commitment may encourage other tech companies to adopt or enhance their own bug bounty programs, fostering a culture of shared responsibility and cybersecurity.

00:28:21:09 – 00:28:44:28
Joseph
Now, I I’ve always been a big proponent of bug bounty programs. I think having outside eyes look at your code and look at your products is the best way to kick the tires. You can have all the internal testing you want, all secure that you want. You can have your developers pour over them. Even your own users go through your applications.

00:28:45:00 – 00:29:10:05
Joseph
But the problem is, is that there tend to be too close to the situation. And a lot of times they’re not going to think outside the box. When you have an outside white hat or gray hat hacker who’s basically trying to poke holes in your software. Internal testing has and I come from a quality assurance background in software from years ago.

00:29:10:07 – 00:29:38:10
Joseph
The problem that I saw when I came to that position was we test it for success. So you have a list of requirements you build your test plans off of that, and you make sure that when I click this button, it does what the requirements say it does. If it does, check the box we’re good to go. And what you tend to have in a situation like that is you have fewer people that are trying to break the software vulnerabilities.

00:29:38:15 – 00:30:14:24
Joseph
While there are bugs, there’s functionality, bugs and stuff like that. You definitely need to catch during that process. And that process is very good for that. The real vulnerabilities that you have to worry about these days are the people that are deliberately trying to break your software for penetration reasons, and if you don’t have that red team, blue team mentality in in your organization, the next best solution to that is having these third party, entities who are inspired by your bug bounty program.

00:30:14:26 – 00:30:42:12
Joseph
It’s been very helpful to a lot of organizations in the past, even big ones. You’ve got IBM, you’ve got Microsoft, obviously, you’ve got Google that are engaging in these. The problem you run into is when you have other companies that are close for for the longest time, Apple resisted bug bounties. They did not want anybody poking holes at their software because they were in a closed ecosystem.

00:30:42:14 – 00:31:04:11
Joseph
When they started cropping up with bugs here in their in their software, they started to wise up and they opened up bug bounty programs too. So I think this is definitely a good thing here. I think Google is very smart. It also may signal a new influx of, applications or services coming from Google with them, expanding like this.

00:31:04:14 – 00:31:39:17
Joseph
So kudos to them for staying ahead of the curve. Next up, TechCrunch is Kyle Wiggers tells us I can’t hack it alone in programing. In a recent interview, IBM CEO Arvind Krishna expressed skepticism about AI’s ability to replace human programmers in the near future. He emphasized that while I can assist in coding tasks, the complexity and nuance of of programing require human expertise.

00:31:39:19 – 00:32:15:16
Joseph
Krishna’s perspective suggests that I will serve as a collaborative tool rather than a replacement for programmers. This collaboration could lead to increased productivity, allowing developers to focus on more strategic aspects of software development. Christa’s viewpoint aligns with the current trend of AI augmenting human capabilities rather than replacing them. This perspective is shared by other industry leaders, highlighting the importance of human oversight in AI driven processes.

00:32:15:18 – 00:32:45:12
Joseph
And I would tend to agree with this. I think AI is great for the things that are redundant, are difficult, not difficult, but tedious to do. And the first thing that comes to mind is, code review, code optimization. One of the things that I’ve always struggled with in every development organization is once a piece of code is written, most developers never want to go back and touch it unless someone points out a problem there.

00:32:45:14 – 00:33:16:12
Joseph
And even if you don’t have out and out bugs that are breaking things as your development technique works, as you continue other projects, you learn new techniques. There’s new technology and it’s worthwhile going back and looking at your existing products and doing code reviews and optimizing them and implementing new standards, new technologies, new coding techniques. And I think this is something where I may be very helpful in trying to identify these things.

00:33:16:15 – 00:33:53:23
Joseph
The intricacies intricacies of programing involve problem solving and critical thinking skills that I, as yet the master. This underscores the need for human programmers to handle complex coding challenges. By automating repetitive coding tasks, I can free up developers to focus on innovation and design, potentially accelerating software development cycles. The integration of AI and code necessitates continuous learning and adaptation among programmers to effectively leverage AI tools.

00:33:53:25 – 00:34:21:18
Joseph
And with anything with AI, we’re about to get in a deep dive. Not really deep dive. More of a historical review of AI as a tool. AI is not a replacement for humans, and the media really has picked up on AI in a distorted manner to make people think that, oh well, I can just do that for you or hey, your students shouldn’t be using AI because they’re going to cheat.

00:34:21:20 – 00:34:39:27
Joseph
None of that’s true. I mean, some of it’s true. Sure, they’re going to use it to cheat, but really, what it is is a useful tool. I mean, you have a hammer. A hammer is great for hammering and nails, but you can do more than that. You can use it for more than just hammering it. Nails, but it’s very good for hammering and nails.

00:34:39:27 – 00:35:11:19
Joseph
So the the trick is finding what the niche is, where it works best, and different types of eyes apply to different types of scenarios. So in this case, here, I agree. I think, I think AI and programing is probably a good thing and certainly is a good thing. I mean, we’ve had it for a while with code completion and stuff like that, but it’s getting to the point now where we can start doing some of those tedious jobs and improve your ability to produce code.

00:35:11:21 – 00:35:36:22
Joseph
So that’s all we had for the news for this week. We’re going to take another quick break. And when we come back, we’re going to start our talk, our deep dive. We’re going to call it it’s not too deep, but it’s deep enough to give you some more information that you probably already have on AI. We’ll be right back.

00:35:36:25 – 00:35:57:04
Narrator
Are you tired of your favorite gaming podcast finishing with a play? Oh no. Well check out no credits rolled or we play the games but rarely finish them. How’s it going, folks? I’m Sam Whalen, your friendly host at No Credits ruled the ultimate gaming podcast where we dish out the latest scoops and reviews on all your beloved video games.

00:35:57:07 – 00:36:19:12
Narrator
Hey, listen, not only that, but we spice things up with some guest interviews and even give you, yes, you a chance to have your say. Tune in every other week for a fresh dose of no credits roll available on all major podcast platforms and hit us up on social media at no credits. Rolled. So why wait? Let’s dive into the gaming world together.

00:36:19:12 – 00:36:25:08
Narrator
We’re finishing games is optional, but the fun is guaranteed.

00:36:25:10 – 00:36:33:11
Narrator
Hey.

00:36:33:13 – 00:37:05:09
Joseph
This is insights into technology, and today we’re talking about AI in our deep dive. Today we’re kicking off our deep dive into artificial intelligence. What it is, how it got here and why it’s reshaping everything from small startups to large scale enterprise systems. In this first segment, we’ll be talking, taking a quick journey through the history of AI, highlighting the milestones that brought us the powerful and deep learning models we see today.

00:37:05:12 – 00:37:34:12
Joseph
We’ll also touch on how this new era of AI impacts commercial and enterprise computing. So AI, understanding the evolution of artificial intelligence, understanding AI is evolution is critical, and most people tend to not pay attention to the history of technology like this. But I think it’s very important to understand where you were. So you know where you’re going.

00:37:34:15 – 00:38:09:16
Joseph
AI is not just a buzzword, but a technological force that shaped industries, redefine business processes and influence daily life. It’s hard to turn on the news these days and not hear something about AI. So how did I evolve from rigid rule based systems to adaptive self-learning models? AI today powers everything from recommendation engines such as Netflix and Amazon to autonomous vehicles and fraud detection in banking.

00:38:09:18 – 00:38:53:18
Joseph
So AI’s evolution really started with. What the work that was done during World War Two, you had the, immersion of of Alan Turing and his idea of machine thinking. One of the earliest sparks in AI inspiration came from Alan Turing. And that’s the Turing test. I’m sure a lot of people have heard about. The Turing test came out of the 1950s, and it was a simple, test, a simple process where he asked a question, can a machine’s answers be indistinguishable from a human’s?

00:38:53:21 – 00:39:23:13
Joseph
And I certainly think we’ve reached that point today, although the reliability and accuracy of those answers is probably something that needs to be, taken into concern as the concept AI is. A concept emerged from the 50s when computer scientists aspired to create machines that could emulate human reasoning and problem solving. Early AI research leaned heavily on symbolic AI.

00:39:23:15 – 00:39:58:09
Joseph
This approach used hand-coded rules and logic to mimic aspects of human thinking. You can think of it like creating a giant decision tree. If this happens, then do that. Symbolic. I was powerful for very specific tasks like playing checkers or solving algebraic algebraic problems, but it struggled with more complex real world scenarios where rules weren’t always black and white.

00:39:58:11 – 00:40:31:01
Joseph
In the 1970s and 80s, expert systems in enterprises began to emerge. Early AI systems were deterministic. They worked on the if then logic, for example, the medical mycin the medical expert system for diagnosing infections in finance. You had systems that were used for loan approvals, and in engineering they were using AI for fault diagnosis in machineries.

00:40:31:04 – 00:41:07:02
Joseph
Now, there obviously were limitations. They couldn’t handle ambitious or evolving situations, and they only worked within predefined knowledge bases. The 1990s with the 2000 was the shift to machine learning. Why the shift? Businesses needed adaptive systems that could learn from new data. Computational advances allowed for statistical models rather than hard coded rules. Some of the key breakthroughs here were neural networks.

00:41:07:05 – 00:41:46:22
Joseph
Although conceptualized in the 1950s, artificial neural networks gained traction with better training algorithms. A neural network is an artificial intelligence. An artificial intelligence is a computational model designed to simulate the way human brain processes information. It consists of, layers of interconnected nodes or neurons. In the case of the brain, these processed data in a hierarchical manner, enabling machines to recognize patterns, make predictions, and learn from the data.

00:41:46:24 – 00:42:19:08
Joseph
Support vector machines or SVMs and decision trees became popular during this time as well, and SVM is a supervised learning algorithm used for classification and regression tasks in machine learning. It’s particularly effective for binary classification problems and works by finding the optimal decision boundary or what they refer to as the hyperplane that best separates data points into different classes.

00:42:19:10 – 00:42:57:20
Joseph
This breakthrough helped to revolutionize fraud detection. Speech recognition. This gave you your early versions of Siri and Google voice search and stock market predictions, as well. Adaptability became a major advantage of machine learning over expert systems. Anomaly detection use cases include fraud detections, banks detecting unusual transactions in cybersecurity with identifying network intrusions, and predictive maintenance, where factories were spotting equipment failures before breakdowns actually occurred.

00:42:57:22 – 00:43:26:05
Joseph
Some real world examples include credit card companies flagging suspicious transactions based on deviations from user behavior rather than predefined fraud rules. So, for instance, if you use a credit card, if you live in new Jersey, for instance, like me, and you use a credit card while you’re on vacation in California, the system might flag that as suspicious and notify you or lock the card or something like that.

00:43:26:07 – 00:43:43:22
Joseph
Basically, using machine learning to better understand your patterns and the ability to, react to outside stimuli at that point in time.

00:43:43:24 – 00:44:24:08
Joseph
Deep learning moved into the 20 tens with a bit of a revolution. So what changed? GPUs provided the computational power needed to train deep neural networks. Neural networks with multiple layers using deep learning began outperforming traditional machine learning models. There was a data explosion. Social media, internet of things, and cloud computing provided vast amounts of training data, which we’re only finding out now, wasn’t really used legally, certainly from a copyright standpoint.

00:44:24:08 – 00:45:03:26
Joseph
And there’s a lot of lawsuits right now. They’re still in progress for some of the key breakthroughs that came out of this period, where image recognition, ResNet and convolutional neural networks, or CNN, has surpassed human performance in some tasks like convolutional neural network is a type of deep learning model specifically designed for image processing and pattern recognition. CNNs are highly effective for tasks involving visual data such as image classification, object detection, and facial recognition.

00:45:03:26 – 00:45:33:10
Joseph
So unlike traditional neural networks, which treat all input equally, CNN use convolutional layers to extract important features from images, allowing the model to detect patterns such as edges, shapes, and textures. Now we’ll talk later all about some of the biases that were built into these systems during their training that today are still an issue, but less of an issue.

00:45:33:12 – 00:46:28:07
Joseph
But there’s certainly much more in the forefront now. People are much more aware of them. Another breakthrough was natural language processing, or NLP. Google’s Bert and OpenAI’s GPT transform chat bots and search engines. You also had autonomous systems, self-driving cars, robots, and AI driven medical diagnostics all emerged from this. So the practical impact of this was that AI, assistive technology and health healthcare started to help with radiology and detecting cancer and medical imaging, algorithmic training and risk assessment, analysis and finance and personalized recommendations from Spotify, Netflix and other entertainments.

00:46:28:09 – 00:47:00:23
Joseph
Large data sets and neural networks quickly became a part of artificial intelligence as well. Deep learning is data hungry. We kind of all know that at this point, which is why we have such a, concern about privacy when it comes to artificial intelligence. Without large data sets, AI struggles, cloud computing mean it possible to store and process vast amounts of data.

00:47:00:26 – 00:47:29:14
Joseph
Some industries that were transformed your autonomous vehicles, Tesla’s fleet learning approach, is one key aspect your voice assistance that you have today. Alexa, Google Assistant I just said her name and it shouldn’t have e-commerce AI driven customer service, chat bots, AI. Everyone loves calling your your local utility off and having to go through an AI prompt, through the phone.

00:47:29:14 – 00:48:01:25
Joseph
I’m sure everyone’s done that. In the past. So some of the future considerations that we have to take into account here include data privacy concerns. We’ve exhaustively talked about these on the podcast recently. How much data should companies collect? Where how should they store it? Where should they store? How much is protected? Should data have an expiration date on it?

00:48:01:27 – 00:48:30:27
Joseph
There’s bias in AI like we had touched on models can inherent biases from training data. How do we ensure fairness? One great example of this was the early work that Microsoft was doing with facial recognition. And they only used a certain segment of the population to train it, that segment being relatively fair skinned. So when they deployed their technology to the field for Microsoft, hello.

00:48:30:29 – 00:49:11:29
Joseph
Or when they were doing work with the Kinect, people of darker skin, African-Americans, Hispanic and so forth. The facial recognition was not picking up on them correctly because it wasn’t trained, with the darker skin tones in mind. Energy consumption, large scale AI training requires significant computational resources. What’s the environmental impact? We already had one data center for AI that was basically built next to its own power plants, and happened to be a coal power plant that was just built because it was cheaper, things like that.

00:49:11:29 – 00:49:46:01
Joseph
Cooling is another issue. I require so much computational power, so much cooling, that they need to be built almost with the consideration of like a nuclear power plant. So you have a cooling resource like a river that you can constantly draw on. So some of the takeaways that that we need to think about here, AI’s evolution has been marked by shifts from rule based logic to data driven learning and deep model multi-layer neural networks.

00:49:46:04 – 00:50:26:16
Joseph
Adaptability is AI’s biggest strength to date so far. Modern systems can learn, predict and act with minimal human intervention, which, as we talked about in an earlier story today, might actually be a problem. But that lack of human intervention, industry disruption is ongoing. AI is still evolving with quantum computing, edge AI, and more. Now, edge AI refers to the deployment of artificial intelligence models directly on edge devices such as smartphones, IoT sensors, drones, and so forth.

00:50:26:19 – 00:50:57:19
Joseph
Rather than relying on the cloud computing. This allows AI to process data locally on the device, reducing latency, improving privacy, and enabling real time decision making. So that’s just a quick look at, the history of AI in, our next, deep dive, we’re going to take a look at AI in the enterprise and how it’s affecting businesses and commercial interests.

00:50:57:21 – 00:51:24:05
Joseph
That was pretty much all we had for the day. Before I do go, I want to quickly have, throw a few shout outs out there. We once again had a very good week on Twitch with a number of, new followers, a couple that, I jotted down here. Mommy, Asmaa. I can’t even pronounce this one.

00:51:24:05 – 00:51:55:02
Joseph
M u m I u p r u Muslim u pew. Maybe. I don’t know. Pronunciation of names was never my strong point. We also had talking to five neural hazar underscore u, Bananarama, underscore and several other people that I unfortunately did not write down for this episode. But I want to thank you all for, following us on Twitch.

00:51:55:05 – 00:52:26:20
Joseph
Before we go, I do want to once again let folks know that if you want to find audio versions of this podcast, insights into technology, you can go to insights WWE that insights into technology.com. You can find us insights into technology anywhere you get a podcast. If you’re interested in getting a video version, you can find both video and audio versions of all of our podcasts up on our website of our main website at Dwarka.

00:52:26:20 – 00:52:47:20
Joseph
Insights into things.com. We are also listed as insights and the things on your podcast sites. As well. I would invite you to give us your feedback, tell us how we’re doing. Give us your opinion. I’d love to hear you know, what you think about some of the stories that we’re talking about, or some of the deep dives here.

00:52:47:22 – 00:53:13:11
Joseph
You can email us your comments and insights into things. Dot com. You can call in, leave a voicemail for us at (856) 403-8788. We do stream five days a week on Twitch, which is where a lot of our followers come from. You can find us at Twitch.tv, slash insights into things where you can get links to all that and more on our official website at that.

00:53:13:14 – 00:53:17:22
Joseph
Insights into things.com. That’s it. Another one in the books.