Removed by mod

It’s helped me further understand that the average person is a moron.

It’s a really good search engine for research topics. And it can do basic coding tasks via a short plain-text description, which can save time. In prototypic fields like science it’s a really good fit, because software does not need to be high quality, often it doesn’t even need to be super fast.

I like the AI in Super Mario Bros when the Koopa Troopas turn around after walking near a ledge.

I generally hate AI, but the things I hate about it are that it’s being controlled by greedy capitalists and forced upon the world as a product, that it’s driving the worst component shortage in electronics history, and that it’s being used to replace human labor. I’ve found genuine uses for it though, and if it weren’t for the capitalists trying to use it to destroy the world, I’d have a mich higher opinion of it. I used it at work to help write code based on a 4000+ page datasheet, it’s pretty good at churning through a lot of documentation to find the relevant parts. It is basically an advanced search engine after all. I also had some good success with it lately (specifically through Google’s AI mode) reverse engineering the RGB chip in some RAM. I had reverse engineered it years ago bit there were still some things I didn’t understand. Uploaded a firmware dump to Google AI and it was able to find some missing details by analyzing or decompiling the binary. I also used it to recover corrupted SPD on some RAM sticks, gave it a dump of the corrupted data and it determined what bytes needed fixing. I think that in the right hands and used as an actual assistant, it can be helpful, but too many people are trying to cede all control to it in the name of profit and ruining everything. I hope that local AI and smaller, more focused models become more common and that the environmental and economicl disaster that is hyperscalers starts to die off.

It made me realise my new coworker is an moron and a fraud in just weeks, not months or years. If he had simply shut his mouth, I would never has suspected him. But thanks to him feeding several internal things into AI, and then bragging about it to my boss, I now know who I won’t be sharing all but the most required info with.

it is strong with any translation tasks.

also gives me a somewhat okay summary before i decide to read something or skip it.

He always calls me Betty and acts like my long-lost pal.

AI is being used to make advances in medicine that could make my life much, much, much better.

Some good uses ive found reliably:

• Plugging in large or complex articles to give a summary or highlights
• Suggesting what I can do to rewrite/improve emails
• Roleplay out performance and HR scenarios
• Create HR documents with accuracy
• Translating documents into other languages. Though when I give these documents to patients I let them know I had to use AI to convert. My workplace won’t convert them for me officially which would be best answer
• Some search inquiries. I found the answers are only as good as how you word your ask and I always check its citations. Rarely wrong for my questions

What it doesn’t do well:

• Meeting Minutes
• Converting my results into documents
• Making document templates or modifying them
• Excel

Though things would also be much easier if my work just provided me with this stuff like documents in the right languages or enough HR they dont have to tell me to initiate my own investigation (They are 6+ personnel short but this is absurd!!!)

Nothing. (Best I can do)

At first I viewed it as another abstraction of search. That does not really work when you look at generative or agents. Pretty soon I see it as a new interface. The old runlevel went 0-5 I think. 3 was comman line and 5 was graphical. now this was changed as it existed before graphics. I see llms as basically a new type of interface like mouse and graphics after terminal. Now it is kinda different in that it runs side by side with the other run levels so I would call it like 3a or 5a or something like that.

I love that it often works as a helpful tool in my dad’s research for things such as generating drafts, summaries or emails. I value the optimization that it brings to malware analysis, especially when it comes to open source communities such as the AUR’s PKGBUILD files. As for my own personal use, I cannot think of a single aspect of my life in which LLMs are present in any way. I am worried - not for my own sake, mind you - that my workplace (a government office) is going to force AI drafting down our throats, which would be great for optimization, but I fear that it will make the dumb people dumber. I know it sounds condescending. I did use some long forgotten website’s AI image generator to generate very specific pornographic images based on Tolkien’s world, which was fun for a few days or weeks. :3

It has democratized coding which I think is cool. People with no coding experience can think of a solution to a problem and have it coded up. It is honestly a very positive future to think of using your own computer to build new automated personal workflows. “Jarvis, make a dashboard of soccer results for the last two weeks, find the median team and display social media feeds for all their players”.

Of course it does mean a lot more junk published software, but the vibe coding personal projects is basically future tech utopia.

Otherwise AI in the general sense is a tool used by researchers and engineers, any benefit there is a good thing.

AI ends up with more problems, but if you look for advantages, then what, useful open source programs with specific necessary functionality have appeared, which without AI would have appeared several years later, if at all, and I like speech-to-text conversion that can work locally, but video generation is a pointless and even harmful technology, in my opinion