Seeking opinions about a failure in marketing technique(s)

[riding_on]

Hi neighbors (and distant neighbors, and those who find Mr. Rogers to be offensive),

As part of the ongoing research my company has been doing, I made available via text messages a large language model (basically my tiny company's version of ChatGPT). I've mentioned this to a number of people, and my company's website is public on the internet, where the information to make use of this free-to-the-customer service can be found. Since I released it approximately 9 months ago, not a single individual has seen fit to even try it out. To my knowledge, it is literally the only version of a large language model that is available for use to even seniors with flip phones and no internet access.

I say "free-to-the-customer" because it costs my company money to maintain its accessibility (flat fee), plus a varying charge for every message it sends or receives. Given the popularity of AI language models and the hype surrounding them, I've been very surprised that its only user is myself. Not even my family or those who assist with my research have tried it, but I admit, I'm very gentle in my prodding of people to help me improve our company services. I don't do things like show pop up windows requesting star ratings and reviews, nor do I offer any kind of "Like" or canned reaction functions.

Please find pasted below an interaction between myself and this program:

QUERY:

Please solve for x in the following equation, and try to explain the steps you took to reach your answer as an addendum to the solution for x: 7x = 89

REPLY:

To solve for x in the equation 7x = 89, we can divide both sides by 7 to isolate x.

X = (89 ÷ 7)

To check if our answer is correct, we can substitute the value we found for x into the original equation to see if it holds true.

7(x) = 7((89 ÷ 7))

7x = 89

89 ÷ 7 = x

So our solution for x is: x = 12.71 (rounded to two decimal places).

The reason we divide both sides by 7 is because we want to get rid of the 7 from the left-hand side of the equation, and leave us with just x. Once we have x isolated, we can perform the division on both sides to solve for x.

Addendum:

To solve for x in the equation 7x = 89, we can follow the steps below:

1. Divide both sides of the equation by 7 (the coefficient of x).

X = (89 ÷ 7)

2. Check if our answer is correct by substituting our solution for x into the original equation to see if it holds true.

If it does, we know that our answer is correct.

For example, if our answer for x is 12.71, we can substitute that value into the original equation:

(7 * 12.71) = 89

This should hold true because we found our answer by dividing both sides by 7.

3. Round our answer to the desired level of precision.

In this case, we rounded to two decimal places.

My website is https://www.srllc.org, if anyone is interested in replying and needs more context to do so responsibly. Thank you for your input, folks.

 

[neptronix]

Please sit down because i'm about to give it to you straight since you asked twice.

I'm the head of product design and development for 4 companies. I also run two businesses. Primarily i work with startups. Been in the computer industry to some degree since '92.

The AI market has been beyond saturated for some years now and people are burned out on the new shiny toy after having AI pushed on them like crazy from big tech.

Yup, getting people to even try your product is a hell of a challenge because the perception of the industry is that the majority of it's products are over-promising/under-delivering.

Also, AI is enshittifying the internet right now.. it's also cutting human-human contact out and creating problems.. so.. the general public perceives it negatively.. you will have a hard time escaping that orbit..


Any AI system is competing with the big boys, who have massive venture capital, and are willing to run their operations at a financial loss for a few more years ( yes, even OpenAI is not making money! )

I run a software development company and wouldn't touch this segment with a 999999999 foot pole unless i had something significantly better than what OpenAI or Anthropic were offering. And if i had that, i would immediately seek venture capital because there's no way you are going to bootstrap this business, pay for the resulting datacenters, and someday make a profit after all the cash bleeding.

What the market wants is AI with better reasoning capabilities than the current state of the art, and higher accuracy. You provide that, and you have a chance of winning.

An AI that is accessible via certain mediums is not solving a problem the market needs solved. it's not a difficult problem to solve, either.

I don't know if you read this but a company whose business model revolved around 'better way to access AI' just died.
Humane AI Pin is a disaster: Founders already want to sell the company

I think you are playing the software development and business game on nightmare difficulty mode, i would consider hitting the up key and selecting, at least, the ultra-violence difficulty level




Also, your website does a poor job of leading me to, and explaining the benefits of the product. If you hired a good marketer or SEO person, they would level your website and write it with from the perspective of human psychology, which is a huge mental shift most programmers are unable to do themselves ( and this is why every Steve Wozniak needs a Steve Jobs, and vice versa )

But before you even remotely consider the above idea, if you want to survive in this market, you need significant differentiation and and improvement on the current state of the art to even get attention.

 

[harrisonpatm]

Also, AI is enshittifying the internet right now

Cory Doctorow deep cut!

 

[riding_on]

Your comments are well received @neptronix, and exactly what I was asking for. A few responses to keep the data stream flowing:

1. I'm able, willing, and foresee providing this product for a few years at a loss to my company. I'm a pseduo non-profit (.org), meaning as soon as I need to claim that I am, I'm ready to at a moment's notice. Venture capital and controlling interest are not at stake with me. My legal filings guarantee perpetual, unilateral control of anything my company has a stake in, but if I'm lacking in providing public benefits, I risk insolvency and bankruptcy.

2. Patent applications are derivative and public for a reason. Pet rocks were a thing, and they weren't a solution to a hard problem. If your statements are true, I assume that investors have had enough of the bullshit (about damn time, if you ask me).

3. I'm comfortable living in nightmare. I've been reported to various authorities as "too dark for my comfort," my entire life. They're so used to it that they don't even bother checking anymore. Literally.

4. Again, thank you for your input.

 

[neptronix]

1. I'm able, willing, and foresee providing this product for a few years at a loss to my company. I'm a pseduo non-profit (.org), meaning as soon as I need to claim that I am, I'm ready to at a moment's notice. Venture capital and controlling interest are not at stake with me. My legal filings guarantee perpetual, unilateral control of anything my company has a stake in, but if I'm lacking in providing public benefits, I risk insolvency and bankruptcy.

Interesting, happy to hear you're a fellow devoted bootstrapper. Same here ( so far )

3. I'm comfortable living in nightmare. I've been reported to various authorities as "too dark for my comfort," my entire life. They're so used to it that they don't even bother checking anymore. Literally.

Good, that's a component of a winning attitude in business!

4. Again, thank you for your input.

You're welcome, thanks for being good at taking what i hope was constructive criticism.

 

[harrisonpatm]

Maybe this isn't the thread for it. But, just as an fyi.

The first version of Paypal's digital wallet/electronic payment system was by a company called Confinity, founded by Max Levchin, Luke Nosek and Peter Thiel. Musk's then-company, X.com, merged with them, called it Paypal, and Musk left as CEO shortly after.

Tesla was founded by Marc Tarpenning and Martin Eberhard in 2003. Musk bought in one year later with money from selling his stake in Paypal. Disagreements between Musk, the two founders, and other principal characters, led to a 2009 lawsuit that resulted in Musk being allowed to call himself a "founder."

Musk did indeed found SpaceX, no lie there.

 

[neptronix]

Musk is a mixed bag but there's a lot to learn from him, and his other imperfect business peers.

I shared that quote because it touches my business spirit. The void told me to tell you, hi.

One of the biggest mistakes he's ever made is talking about politics. The social and economical repercussions to him and his companies have been huge. I have learned from his mistake and decided to, for the most part, shut the f*** up about politics myself

 

[riding_on]

Musk is a mixed bag but there's a lot to learn from him, and his other imperfect business peers.

I shared that quote because it touches my business spirit. The void told me to tell you, hi.

One of the biggest mistakes he's ever made is talking about politics. The social and economical repercussions to him and his companies have been huge. I have learned from his mistake and decided to, for the most part, shut the f*** up about politics myself

I'll talk about anything I feel responsible talking about, and I'll do so with anybody I feel has an interest in the topic, or has a need to know the things that I do.

I refuse to comment on specific legal cases, however. I've come to understand that the law is far too nuanced for outsiders to understand -- it's best left to those who were involved to begin with, and nobody else. Here's hoping that someday I appear slightly less pretentious for comparing myself to Elon Musk

 

[lnanek]

At least in the indie game dev world, we try to build our games with a good viral coefficient, and even then we have to pay a good $80k in ad spend just to get in the top listings of the app stores briefly enough to pickup any users whatsoever. After that, if it's well designed, it can grow organically.

E.g. instead of a free product, what if it was only free for the first month, then required you to refer a friend to keep using it? Or what if it had the ability to do group conversations, so if you asked or to plan a vacation, users would naturally add another person? That's increasing the viral coefficient.

That said, that site, doesn't even have a good conversion funnel going, so I think ad spend would not be valuable at this point. I had to follow the link, then saw a bunch of research projects, and only after digging through a menu did I eventually get offered free texting to the chat bot. But the number wasn't even a hyperlink, so I couldn't click it to copy it or open it or anything. At that point I gave up since I already have apps on my phone that make this easier like Snapchat's AI buddy.

If you consider everyone landing on the page to be the top of the funnel at 100%, then the people who dig through the menu may be like 10%, then the ones that actually manually copy the number and text might be 1%. So you can see how your user funnel was big at the start then tiny at the end, right? If you are paying a dollar per user sent to your site in ads, you just changed that to paying $100 dollars per user of your product due to a poor conversion funnel.

You could have had the the texting number be the first thing on the first page without scrolling or menus, and could have had a live chat interface, link to an app, or at least a hyperlink to text, or at least a link to copy the number to the clipboard. Then your conversion funnel would be better and cost of user acquisition might only be $10 per user instead $100.

 

[riding_on]

At least in the indie game dev world, we try to build our games with a good viral coefficient, and even then we have to pay a good $80k in ad spend just to get in the top listings of the app stores briefly enough to pickup any users whatsoever. After that, if it's well designed, it can grow organically.

E.g. instead of a free product, what if it was only free for the first month, then required you to refer a friend to keep using it? Or what if it had the ability to do group conversations, so if you asked or to plan a vacation, users would naturally add another person? That's increasing the viral coefficient.

That said, that site, doesn't even have a good conversion funnel going, so I think ad spend would not be valuable at this point. I had to follow the link, then saw a bunch of research projects, and only after digging through a menu did I eventually get offered free texting to the chat bot. But the number wasn't even a hyperlink, so I couldn't click it to copy it or open it or anything. At that point I gave up since I already have apps on my phone that make this easier like Snapchat's AI buddy.

If you consider everyone landing on the page to be the top of the funnel at 100%, then the people who dig through the menu may be like 10%, then the ones that actually manually copy the number and text might be 1%. So you can see how your user funnel was big at the start then tiny at the end, right? If you are paying a dollar per user sent to your site in ads, you just changed that to paying $100 dollars per user of your product due to a poor conversion funnel.

You could have had the the texting number be the first thing on the first page without scrolling or menus, and could have had a live chat interface, link to an app, or at least a hyperlink to text, or at least a link to copy the number to the clipboard. Then your conversion funnel would be better and cost of user acquisition might only be $10 per user instead $100.

Invaluable information. You have my sincere gratitude. I'll also say that you have an excellent point: even by my standards, it should be a hyperlink. I'll fix that tomorrow.

I'm surprised that nobody has commented on the way I literally deferred to the only people I trust when describing my previous projects (Federal agents and government employees). I just trusted them and quoted their words. If you want to see how I describe the things I do, you're welcome to waste half an hour perusing The Lich Project - Home -- that's from 2005.

 

[Empagstyr]

I ran into a similar issue last year—loads of effort, barely any traction. What helped me was tweaking the copy to speak more like a real person, less like a sales pitch. Also, focusing more on bing seo brought in some decent organic traffic, especially from people not already swimming in Google results. It didn’t blow up, but it got the right eyes on my stuff.

 

[hillslayer@]

Musk is a mixed bag but there's a lot to learn from him, and his other imperfect business peers.

I shared that quote because it touches my business spirit. The void told me to tell you, hi.

One of the biggest mistakes he's ever made is talking about politics. The social and economical repercussions to him and his companies have been huge. I have learned from his mistake and decided to, for the most part, shut the f*** up about politics myself

Talking politics is only a problem when you're a Nazi.

 

[harrisonpatm]

Talking politics is only a problem when you're a Nazi.

We're just coming off issues with unnecessarily inflammatory comments. Let's keep this so-far-appropriate off-topic thread in the clear so we don't have to add this one too the garbage bin too.

 

[hillslayer@]

Just the truth bro. Shouldn't be anything wrong with facts

 

[neptronix]

Sorry, i call foul too. You know this is a super politically divisive comment and is the exact kinda tone that that starts flamewars on this forum with regularity, especially since political tensions are extremely high.

Please read: Rules of the off topic section.

I would be very mad if you disrupted an active thread and it turned into a flamewar.
But this is an inactive thread about some guy's business. So i'm just here to dump some cold water on..

 

[hillslayer@]

If we can't speak the truth I will delete this forum.

 

[neptronix]

It's not about that at all. I don't even disagree with you. I personally have a negative view of musk. But you should note that my post predates his involvement in government.

You can speak what you believe is the truth, that's no problem.

I just need you to move a few notches up on the debate pyramid and not be laying the most excellent kind of political flamewar kindling on some guy's thread about his business.

No matter where you stand on the political spectrum, this standard of behavior applies to you. It also applies to me. Nobody gets an exception.

 

[hillslayer@]

I wasn't calling anybody on the forum name's he should be open to whatever we want to say he is.

 

[neptronix]

Yes, but that's still political flamewar bait on some guy's thread which is not related to that.

Glad we have no bites

 

[hillslayer@]

I responding to your commit on why musks businesses are tanking.

 

[neptronix]

It's fine bro.
His wound are primarily self inflicted. It's a pretty good lesson to any business owner watching him, of what not to do!

 

[riding_on]

It's fine bro.
His wound are primarily self inflicted. It's a pretty good lesson to any business owner watching him, of what not to do!

The truth is that AI is progressing at such a dizzying speed these days, that I've all but abandoned trying to keep up with my competition. Just a few years ago I could implement most of the whitepapers I could find (the useful ones). These days I'm lucky if I don't embarrass myself by having to ask colleagues questions like "WTF does your team mean when they use that word, because to me it's not even a software thing?"

In short: get a good team and maintain your ties with them. I don't regret trying to build something before I had the funds to hire a team to work with me, but in hindsight it was a terrible dead end for my resume.

As for Musk? Meh, I couldn't care less. His ego will survive only having more money than God, as opposed to all that money plus constant success. The rich have it so very hard ... poor unemployable bastards.

 

[neptronix]

This field is insane to get into. Nobody has a moat. Asian open source is verging on achieving technological domination. AI integrators still consistently do a bad job. The technology has barely exited an infantile stage. Trillions of investor cash burned and nobody can claim a net profit yet. Current state of the art is still trained on copyrighted works and many big companies are in court now over that. And the electrical consumption problem continues to get worse. It's a mess.

There's a big number of people who can build AI based on whitepapers at this point.
One of my personal rules is that if i can't do better than average, forget it - focus on a better idea.

Conversely if you have a unique idea, and both you, and others, agree that it is superior.. that tends to be a green flag.
I personally won't build anything until i have that and a number of other checkboxes ticked. But that's the most important one.

But... you could make some bucks integrating AI in a way that works. This requires that you, as a technologist, have good taste and high standards that others don't have. The world is full of broken or bad AI implementations that either need removing or improving.

That's a business segment i've thought about getting into, as soon as the cost for hardware goes substantially down and we are done with the early part of the S curve of adoption.

I think we are just barely entering the 'early majority' phase with this technology.



You have experience with AI integration and therefore have a leg up on the next stage versus technologists like me who have zero experience to date.

I leave you with a quote from an absolute stud of a businessman-engineer:

 

[Geinvedy]

Feels like the real snag might be unclear value for the user. A text-based LLM sounds cool, but people usually need a quick reason to try something new, especially over SMS. I’ve found that looking at how folks in B2B Digital Marketing frame benefits helps a lot, since they tend to focus on simple outcomes instead of fancy features. Even small tweaks in phrasing can shift response rates more than expected.