the-ai-problem/

[utterances-bot]

The Ai Problem

Having Fun with Technology

https://christitus.com/the-ai-problem/

[Konkonfaux]

I haven't used AI intentionally so far. Well, I talked with Copilot a bit. It seems like increasingly the AI will be part of normal usage of software, and there won't be an opt out.

[hernanjbarraza]

I don't use AI at all. I can't seem to find any practical use for it, and am sure many people say "I use AI a lot" just to appear modern.

[erickleung]

This is a 2 edge sword problem. As a Chinese speaking users, we are happy to have hand written recolonization technology developed last twenty years. I think they are the most successful tools or product or main reason for such tools are fund and being developed. However, it is still a product by human, it is not in the same level which is born in a human mind. It is still not able to realize wrong idea or pattern, and correct it by learning from failure, and not mention the 'air' that is unspoken communications of media such as existing in Japan culture or society...

[rron298]

I find the so called AI like Leo uses in Brave browser to be pretty much Wikipedia regurgitators. It is annoying that when it has no data, it won't say I Don't Know, but instead spews useless common knowledge or even propaganda in some cases. Many AI of this type don't "learn" from interaction but are basically a somewhat useless browser "enhancement" to give basic subject overviews at best. It has limited access to data in that it cannot open links on a referenced web page. It lacks even a decent inference ability. It often ignores questions or parts of a conversation and offers unsolicited subject tangents. By the time you drag it's inept carcass back on subject track ,you realize what a waste of time it is. X -AI has it's own issues beginning with it's first choice of data reference. That being the X platform users posted data. I am currently trying an uncensored Ollama LLM... I haven't used any that I trust for accuracy yet.

[kairopup]

Wow.

…

________________________________ From: HyTeqk ***@***.***> Sent: Thursday, July 31, 2025 3:15 PM To: ChrisTitusTech/website ***@***.***> Cc: kairopup ***@***.***>; Comment ***@***.***> Subject: Re: [ChrisTitusTech/website] the-ai-problem/ (Issue #242) [https://avatars.githubusercontent.com/u/36812710?s=20&v=4]rron298 left a comment (ChrisTitusTech/website#242)<#242 (comment)> I find the so called AI like Leo uses in Brave browser to be pretty much Wikipedia regurgitators. It is annoying that when it has no data, it won't say I Don't Know, but instead spews useless common knowledge or even propaganda in some cases. Many AI of this type don't "learn" from interaction but are basically a somewhat useless browser "enhancement" to give basic subject overviews at best. It has limited access to data in that it cannot open links on a referenced web page. It lacks even a decent inference ability. It often ignores questions or parts of a conversation and offers unsolicited subject tangents. By the time you drag it's inept carcass back on subject track ,you realize what a waste of time it is. X -AI has it's own issues beginning with it's first choice of data reference. That being the X platform users posted data. I am currently trying an uncensored Ollama LLM... I haven't used any that I trust for accuracy yet. — Reply to this email directly, view it on GitHub<#242 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/BULFG3FCZ5LBJJID3ORHGIT3LJ2PTAVCNFSM6AAAAACAR4SF4CVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTCNBRGIYTSOJVHE>. You are receiving this because you commented.Message ID: ***@***.***>

[gmanbaron74]

This bubble is going to go pop, very soon.

[BMS1011]

I've used AI for tasks such as troubleshooting code (on personal projects) and organizing a bunch of raw data in .csv files that would've taken me hours or days to do myself. AI/LLMs are very useful in that regard.
But I have the same reservations many others have expressed about the sheer quantity of private information many near-essential or ubiquitous services like social media and search engines are gathering on us in the name of improving their proprietary AI models.
It vastly increases the damage that a data breach can cause. Or worse if the company itself is secretly using our data for purposes the majority of the population would disagree with if made public.

[RobertFrancisBeck]

If ya think 'AI' is a problem, then you probably have some problems to be to be concerned with. I embrace 'AI' for myself, and for the rest of Humanity! Is there a problem?

Semantic Shenanigans