1

With the hype on the internet and TV, there’s an obvious question I haven’t heard asked. Is IT even using AI yet?

AI has the potential to open many security and integrity issues with data. Aside from that, AI has the potential to take over many repetitive tasks such as troubleshooting, updates, as well as configurations!

What are your thoughts as an IT professional, will IT adopt AI to eliminate some tasks? What will AI contribute or take away from the way we provide technical support? Let us know below!

@Intel_Corporation ​

#IAmIntel

18 Spice ups
2

Most of what we deal with is either proprietary company data or intellectual property, so we don’t paste that into ChatGPT and ask what’s wrong with our code or ask for a pivot table of our latest sales figures that marketing asked for.

On the other hand, we’re vigorously looking for ways to use it to make our products better because everyone in marketing wants to say we use Artificial Intelligence, but thinks that Generative AI is somehow General AI and is applicable to every business problem.

5 Spice ups
3

We’re using AI to improve network security, but there are real concerns feeding proprietary & confidential data to systems like ChatGPT.

KnowBe4: Employees Are Feeding Sensitive Biz Data to ChatGPT, Raising Security Fears

As for hands-on IT, until Boston Dynamics or someone else builds robots that can fix things just by walking into the room, our jobs are safe.

cec42144-9720-4904-81ac-80c2bc9064f2-402209444_745940680895081_8155208406846877512_n.jpg
cec42144-9720-4904-81ac-80c2bc9064f2-402209444_745940680895081_8155208406846877512_n.jpg800×478 53.4 KB
9 Spice ups
4

I’ve used AI to generate a few complex migration plans to check and see if I missed anything in my human-generated creations.

Also ran a few oddball error messages past it and obtained actionable answers without having to contact manufacturer support.

I only use ChatGPT from non-work systems and do not include any company proprietary info.

4 Spice ups
5

I’ve been a one-man shop for many years, and major confession: I’m not an expert on everything. RG, when an esxi host throws a weird message at me, I paste it into ChatGPT and ask for complete detailed steps to correct it. (Want to make an awesome product? Add a chip/daemon/service/TSR that watches for error messages and automatically uses gAI to either notify me with the steps to fix or automatically fixes it for me.)

Now that I’m in a developer role, I totally ask ChatGPT to write initial frameworks for me. Very seldomly do I paste code in, and when I do, I make sure to remove or change any api keys, database names, connection strings, etc. I view it like posting a question at StackOverflow (but without the smug answers). I also have GitHub CoPilot installed in Visual Studio and I am still amazed when I add a comment and a 30 line block of code appears doing exactly what I want. Still, even though they say my code is discarded after submission, that part of my brain that doesn’t trust Facebook/X/Google cries a little when this happens.

3 Spice ups
6

I will agree that with some things AI can be good, otherwise it is a threat to humanity without question. If AI is not controlled properly, I can envision Skynet becoming a real thing. There was a reason why a lot of big tech leaders called a meeting months ago to discuss AI as it was seen as becoming a threat.

We cannot ignore the potential of bad use of AI, because AI will learn bad things and then it will be out of control. Just my thoughts as I do not use ChatGPT because I am old school figure it out for myself type for the last 23 years in IT. It hasn’t failed me yet.

4 Spice ups
7

I am using it more and more for quick code snippets. It seems to be very good at that.

I will sometimes use it as a “hail Mary” when I can’t find the answer through traditional searching. Though, this almost never results in a worthwhile answer.

AI also seems to be pretty good at summarizing or de-jargoning various topics.

For example, I recently had a medical procedure and I pasted the incomprehensible doctor’s notes into the chat and it gave me a very clear understanding of what it actually said.

1 Spice up
8

We can’t control AI. What we need to do is embed it with human empathy. Take a look at Scott Jucha’s Silver Ships series for a really good example of how this could look.

2 Spice ups
9

mostly for reports …i just type quick and dirty what i want to say and then asking it to make it more formal …saves me a bit of time

also i know the company that does audit the security logs using ai to detect potential issues .

another awesome use i found for it is asking it to behave exactly like a specific linux kernel and in case the commands i give it do not work explain me why and give me possible solutions …since linux is not my strong point … it helped me tremendously

10

@mikeober ​ Great plan!

AI presents a huge risk for employees that may want to use it to help with their productivity. But as with anything else we must stress security to our organizations!

Thank you for all of your great replies!

@Intel_Corporation ​

#IAmIntel

11

Hasn’t anyone seen Terminator or Battlestar Galactica? Spoiler alert, things don’t end well for humanity.

NO AI for me please.

Yes this is where I will age out of IT I’m sure, I just don’t buy into AI doing my job for me, **** that ****!

12

@markrogalski ​ I understand! AI has the potential to eliminate many tasks not just in IT but throughout the office. For now, I think most jobs are safe.

13

Silly little python scripts. For example a QT GUI that will open a text file with multiple lines that will then wait 5 seconds for the user to select focus and start pasting line by line with a enter key press at the end of each line.

14

I’ve used ChatGPT once to generate an Excel VLOOKUP statement that looks at 2 different columns of a reference sheet to bring back the correct data point. I was surprised at how friendly and conversational it was. The statement works, too. I tried Google first and couldn’t find anything useful.

15

ChatGPT/AI is great for creating documentation, on relatively known subjects.

Like: How does a Non-Profit capitalize on Social Media?

In seconds you have a decent document and with a little smoozing you got a good detailed document, based on facts.

16

Open the pod bay doors, HAL.

I’m sorry, Dave. I’m afraid I can’t do that…

1 Spice up
17

I agree with @mikeober ​ 's graphic. I’d have to say in my day to day, I would rather better search results. I hate having to search the same error with different variations just to get something that fixes it.

18

@randomparts ​ I know what you mean. Searching through blogs, sites, and chat rooms is time consuming!

@Intel_Corporation ​

#IAmIntel

19

and you left out Dune!

Oh, to answer OP, not really for work based on privacy concerns.

Addressing whether AI is good or bad, well it’s gonna be like any IT sys admin tool, ie in the wrong hands mild to very dangerous. Gotta remember there are bad actors out there and even well-intentioned bumblers. It won’t have to even be autonomous before it’s dangerous.

So when will AI have its “flying planes into the tower” moment? AI will absolutely become a threat.
And eventually, AI is meant to even be autonomous - no way to necessarily control it after it gets there (maybe still a long ways away).

@mikeober @jack-intel