Labels

AI DOES NOT NEED TO BE IN EVERYTHING!

While I endorse the use of artificial intelligence (AI), injecting it into everything is stupid despite having "intelligence" in the name. I want to be clear that there are real world benefits to AI to reduce friction for accomplishing a task but that also comes with significant compromises you are probably already aware of and do not need me to mention.

These days we can't turn a corner without AI being mentioned as some sort of solution for everything. It is annoying, pervasive even, that feels more like marketing straddling propaganda in order to have the population adopt the technology willingly and at worse by force. The ubiquity of "AI" reached a point of exhaustion where its proliferation encountered resistance from people (My blog post, not written with AI, is not excluded from this conversation either).

With that being said, drawing attention to areas where AI can be beneficial or detrimental can maybe help steer the technology to where it is best suited rather than slapping it on everything as a panacea, band-aid, or solution. Yes, it is a "Swiss Army Knife" but sometimes simpler solutions can outweigh a complex one offered by AI. For example, going out and experiencing nature without technology is far better for your health than consulting AI in the corner of some dark dank basement. A hammer can drive a nail better than using the handle of a drill.

All jokes aside, rather than inundate you with marketing jargon of infinite possibilities and apocalyptic melodrama that these AI CEOs harp on, we can look at actual scenarios that some are familiar with. This is a nonpolitical view of AI from personal experience and experiences from others.

AI Image from Pixabay1.

DISCLAIMER & DISCLOSURE

I do use AI but at a limited capacity. I do not have and will not use AI agents in my main development environments but may create a containerized setup that does not directly interact with my main development environments. I specifically add some friction to my workflow as a general effort to keep skills "warm". For example, unless otherwise noted, I write the blog posts without AI. Further downstream of my process, I use it (with Google Docs), to some degree, to get a final draft. From time to time there will be mistakes. I generally find amusement seeing how much of what I write needs to be revised where I also include a revision history for each blog post.

Another key point is that some vendors have included AI into the products/services they provide. These are often opt-in by default and in many cases require manually disabling them. For example, Google's release of Chrome includes an opt-in miniaturized version of Gemini for the browser that you need to manually disable and remove, I wrote about it in an earlier blog post, "How-To: Google Chrome: Toggle ON/OFF Built-In AI"2.

Before the AI craze took place and became a societal lexicon, "Siri" for the iphone is an opt-in feature that is enabled by default. Any time you install an application, Siri is enabled and you have to disable it manually for each application. Whether someone finds it useful really depends on a person's use case. If you are not aware of it, you are now.

WHERE IS AI?

At this point, you can't avoid AI. Rather than do a list of where we've encountered it, the best way to determine use cases is by looking at where AI has been implemented to some degree. The examples listed below are in no way exhaustive but helps illustrate where it is effective or where it falls short.

WHERE AI IS MAKING THE SERVICE WORSE:

Amazon

One of the most overlooked aspects of why Amazon is the juggernaut of today is the company's search engine to find products. Over the years, I've lauded that the search engine does an amazing job (and continue to do so) of finding obscure items, more specifically when a person needs to describe the item when you don't know the name of that specific thing. Amazon's search engine more often than not can come extremely close if not the exact item you were looking for. Over the years I've noticed it getting better, before AI adoption.

When Amazon introduced AI to search, first with Rufus, now Alexa, it just gets in the way of the experience forcing users to navigate to the AI graphical user interface (GUI). The search query already provided the necessary information, AI just added a step to repeat the same process of the query and try to force navigation through it. To this day, Alexa AI is super annoying where it is more bloat than beneficial.

Yahoo Mail

The same annoyance can be said with email services3. People, who I've anonymously interviewed, have expressed that the performance for Yahoo! mail for the browser and mobile applications are worse after including AI as a feature. More emails get flagged as spam and doing a search through the emails just hangs up where no results get shown or are irrelevant. Many have found that AI summaries of emails are not that useful, at least at the moment. These are just some of the common gripes for emails specifically. Many have expressed that prior to AI, these grievances did NOT exist.

Summary of Amazon And Yahoo

Fuzzy search is sufficient for search, with engines such as solr and lunr being some of the most common. These are very well established technologies specifically to query text for a large data set. Considering these companies have so much stake in AI's growth, it wouldn't be out of the question to place it on another service that doesn't need it. Having a Ferrari is nice, but if you're friend lives next door, and you want to visit that friend, it might just be better to walk to his/her house rather than take the car.

Controversial Topics Exclusion

Some topics can't be researched with AI. It is not difficult to figure out that there are levers to AI and they contain restrictions that will prevent you from getting particularly sensitive information. Whether you consider this right or wrong can be debated. As the use of AI expands this will become even more common and at worse maybe even alter the information itself rather than outright preventing the search.

Military Miscalculation

Reports of USA military AI misidentified a Chinese vessel possibly containing weapons, exacerbating the already growing tensions between both nations. Despite this mishap and an anticipating of more of them, the USA military is fully committed to adopting and "experimenting" uses with AI4. Obviously this comes at the financial expense and the expense of injury to people. One can only determine what sort of oversight is being handled in order to validate whether AI is making accurate assessments of situations.

Local AI Tools Stealing Your Information

Let's be real... AI companies steal information from its users, stored in their servers, and what is available on the internet. That data is training their AI models, digests them, and serve it back to us5.

As aforementioned earlier in my disclaimer, I don't host localized AI. A great example of why is because of situations like this, back in July 2026, a developer encountered Grok CLI spaceX/xAI uploading the contents of his whole home directory to Grok servers 6. Who would have thought?!

The irony of the whole situation is that we are "supposed to" trust these companies that want us to adopt the technology. However, examples of egregious behavior goes un-penalized because they are considered an "accident" and the blame is set on us. And in some regards, you can consider software content coming from "trusted companies" as malware that society is encouraging you to install on your device, sometimes whether you like it or not.

RESOURCE CONSUMPTION

We should also take note that AI uses more computer resources compared to general compute. This alone makes adopting AI to everything a little questionable if more resources are required for a task that could be accomplished faster without it. Fundamentally, what this all means is that all this AI hoopla requires an incredible amount of electricity.

Summary of AI issues

As prefaced earlier it is not perfect. A person acquiring information that was queried by AI needs to be doubly careful that the information provided is legitimate and correct. There are cases where AI can hallucinate or get information completely wrong. In some regards it is almost like shaking an 8 ball and waiting for a message to appear in the small window.

The military and the Grok examples should put AI shills in the back corner. These are just pure examples of egregious behavior and little to no oversight over the use of AI on different applications. These are HUGE failures by any measure and can make any individual or people victims without any repercussions.

There are a number of great uses for AI, just not on everything. We should really take a careful look at where it is being adopted...

WHERE AI IS MAKING THE SERVICE BETTER:

Optimized Search Summaries

Search engine's with AI for the web is awesome, albeit, at the expense of websites being hidden behind the summaries from the AI search queries. It is not perfect but speeds up the process for researching topics, where a good summary and examples are typically provided. I find that more often than not, I will click on the referenced material that AI summarized because I need to delve a little deeper into a particular topic and sometimes the summaries don't give quite enough information. This alone speeds up building an understanding of material that is being researched.

Code Generation

As a programmer, I am not against people creating "stuff" with AI. I highly encourage it. Most of the content that AI generates come from open source resources available on the internet. In many use cases, someone without a technical knowledge will be able to create what they envisioned. This is great for entrepreneurs with limited resources. Someone who is more technical will extend their abilities further because of having the base knowledge of particular topics. So, it is a win-win.

CODE EXAMPLE

Just because AI can spit out code for you, it doesn't mean the answer that it gives will be correct. Having knowledge of the technology may be required if you happen to run into a hiccup. For example, take this code that was provided by AI written in rust using an external crate called, headless_chrome. The code will not compile despite it being syntactically correct.

  • Bonus points if you can spot where the issue(s) occur7? Write your answer in the comment section.
use headless_chrome::{Browser, LaunchOptions, protocol::page::ScreenshotFormat};
use std::error::Error;

fn main() -> Result<(), Box<dyn Error>> {
    let launch_options = LaunchOptions::default_builder()
        .headless(true)
        .browser_path("/path/to/browser") 
        .build()?;

    // Launch the browser
    let browser = Browser::new(launch_options)?;

    // Create a new tab and navigate
    let tab = browser.new_tab()?;
    tab.navigate_to("https://www.example.com")?;
    tab.wait_until_navigated()?;

    // Example: Capture screenshot
    let screenshot_data = tab.capture_screenshot(ScreenshotFormat::Png, None, None, true)?;
    
    // Save screenshot (omitted for brevity)
    // std::fs::write("some_screenshot.png", screenshot_data)?;

    Ok(())
}

Art Creation (Movies, Pictures, Music, etc)

I am definitely going to divide people on this. I support using AI to create art. Here is my reason:

ABSOLUTELY, NOTHING IS STOPPING ANYONE FROM CREATING ART. NOTHING IS STOPPING YOU FROM CREATING. Creating is a mindset; AI is a tool. AI is an advanced version of a paint brush. Whining about AI replacing this and that is just an excuse to avoid the actual process of DOING. You have to remember that people that use AI to make content are the people who are ahead of you because they took the initiative to ACTUALLY MAKE SOMETHING. You complaining is your version of "cope" and you want to replace the responsibility on something else, other than yourself. You DON'T have to use AI if you don't want to. The biggest hurdle we have to overcome is our excuses we make for ourselves.

For instance, I love writing. In my disclaimer I talk about at what capacity I use AI. Could AI write stuff for me? Absolutely. Here is the crutch, I love writing, so, I write. I don't give a shit if people are not reading my content. I love the process of writing, the same way I love programming because they are both creative processes. If anything, I feel like AI has in some way helped me enjoy the process more because it is an aid for these things and it is a very advanced tool at our disposal. The content that I create is for me first, that as an outcome also serves to benefit others.

We have to take into consideration that technology over the years have lowered barriers of entry to get into basically any field of interest any person across the world can pursue. For example, if you wanted to make a high production movie say 30 years ago, you needed Hollywood or advanced tools that cost hundreds, if not hundreds of thousands of dollars, to produce a movie. Now, you could literally make a hit with just your phone and have an audience across the world in an instant. The gatekeepers to these fields are now only a smaller section of those industries where anyone can establish themselves and draw an audience just by creating content that people like.

  • Fight me on this in the comments section.

Summary of AI benefits

The simple answer is that AI is inclusive. The barriers for entry in just about any field has been lowered even more. Books were a way to persistently store thoughts and information. Once you learned to read (and write), the information was available to you at your disposal. The internet did the same, just at scale. AI, furthers that even more.

CONCLUSION

Regardless, of the many opinions of the AI technology, it is here to stay and will continue to proliferate as time passes. Talking about artificial intelligence (AI) is a divisive topic. There are going to be mixed views about it. There are clear benefits of using AI and clear problems in the same vane. Every person will come into contact with AI to some capacity but our uses will vary from person to person.

For me I support it, however, I use it at a limited capacity for many different reasons as described throughout the blog posts. In many regards, I hope that my reservations about the technology can help encourage people to be more conscientious about adopting it and maybe even put a few barriers in front of it as a long-term protective measure.

We can not ignore the dangers of AI but at the same time can't ignore the many benefits that it offers. Finding where AI is best suited will be a touch and go process. Here are some clear benefits and negatives of AI:

BENEFITS

  • If there is a skill that needs to be acquired, AI is good for summarizing and setting up guardrails.
  • Speeding up processes including creative ones.

NEGATIVES

  • AI hallucinates.
  • Over use on applications that did not need it.
  • Propaganda.
  • Misuse.
  • Intellectual property theft.

Probably the thing I despise the most about AI is how the CEOs of companies talk about how dangerous it is to humankind and expect people to just accept it as a benefit. It is this sort of mindset, and talking from both sides of their mouths, that has really divided people on AI. Without getting political, this sort of rhetoric is mostly occurring in "Western" led AI companies and to put it mildly, that this sort of talk is a little disconcerting because we recognize that these talking points could very well be part of the intent and ambitions. I would argue these are prerequisites but let's not jump into the the conspiracy rabbit hole...

Everything comes down to responsibility. There is no "right" answer. On the one hand, we are at the edge of a slippery slope, but on the other, we are going to come across more breakthroughs. A tool in the right hands builds while the other destroys.

Even when we know that AI can be an immense resource for information, summarizing, learning, etc, for whatever reason we have, we must take considerable care in recognizing that not everything AI can show us is real or fact, despite AI making it almost indistinguishable. Many times the information provided can either be incomplete, misinterpreted, or completely fabricated. Some people take whatever AI spits out as doctrine and solely rely on AI as an external source as a source of truth rather than make the effort to contest it, even against your own beliefs. You can't let AI rob you of your agency and intellect even though it is so much easier to outsource8 thinking to someone or something else, in this case, AI. A person should not export/outsource their thinking process to artificial intelligence and should not double down on it the same way that the many nations did with manufacturing. Minimize dependency but leverage how it can improve existing systems. In other words, we should try to avoid making future society turn into the movie, "Idiocracy"9.

That is part of the reason why I restarted my blog because I recognize that there could be pieces of information that people are gathering from AI that could have some missing pieces. I have every intent for AI to scrape information from my blog posts, unless otherwise noted, in the hopes that it reaches people and helps clarify certain topics or even build on top of the ones that already exists.

This might sound a lot like I am talking about Doom and Gloom but in actuality, I honestly feel like the future will look more like "Star Trek Next Generation" or "Wally" rather than the "Terminator". If I haven't quite illustrated that I have mixed views about AI yet, my basic assessment is that we should proceed with caution and at the same time, not condemn AI all together.

  • Are there specific areas where AI has been beneficial to you?
  • Do you see other areas where AI has made negative impacts?
  • Where do you see AI ten years from now?
VERSIONS

2026v.0.1.1: 4Oct26: Wording cleanup

2026v.0.1.0: init10

REFERENCES

  1. pixabay:BrianPenny BrianPenny ↩

  2. How-To: Google Chrome: Toggle ON/OFF Built-In AI ↩

  3. These are second hand accounts expressed by grievances from users I have spoken with. ↩

  4. Exclusive: US Military had close call after using AI for false intelligence report, sources say: We should also take this reporting with a grain of salt and maybe a shot of penicillan, considering that this news source, most often than not tend to lean on opposing right leaning administrations. The same could be said with left leaning administrations. Regardless of your opinion, if the reporting is true, this poses a dangerous precedence of AI misrepresenting information especially for military use. With that being said, we should also practice some caution when a report has, "sources say". This is not an endorsement for against political views but rather an observation of AI adoption ↩

  5. If you've ever read through a TOS, then you would be familiar with this. It is not inherently negative but should taken into consideration when uploading information about yourself on a remote server that you don't own. ↩

  6. Hacker News: Grok CLI uploaded the whole home directory to GCS ↩

  7. I've already resolved this issue for myself, so I am not trying to get someone to solve it for me. I am using this as an example to describe that solving this issue will be much easier if you know the underlying technology you are working with. headless_chrome is accessing Chrome and chrome features to browse website without the browser GUI. ↩

  8. I can't remember which YouTube video I watched that described how people are outsourcing their thinking and it is a very good interpretation of the phenomenon. Whether or not that person thought that all on their own, I am not sure but it is a novel way to describe the use of AI. ↩

  9. IMDB: Idiocracy ↩

  10. blog.avrstory.net's development release page. ↩

Labels:

Comments

Post a Comment