3 ms·

But eventually people will catch up you can basically create a working product alone with the help of AI. My prediction is that this will lead to a margin free

But eventually people will catch up you can basically create a working product alone with the help of AI. My prediction is that this will lead to a margin free-fall for many software products where the main moat is the software itself. And a lot of SaaS companies will also become redundant when the AI can code up a tailored solution in an hour for free.

CharlieDigital 9mo ago
You are discounting sales, marketing, and branding. Take drop shipping for example: anyone can do this, but the successful ones are those that know how to brand and market the product well. Not to mention having the right mindset for startups and building a business. The code and product is maybe only 20% of the story.
- margorczynski 9mo ago
  I'm not. That edge eventually converges to 0 when you have 10+ competitors that offer the same for 10x less money. If you don't have some kind of cult following like Apple eventually you'll get margin-squeezed till death and all that marketing, sales, etc. will get cut down to stay afloat. Of course all of the above is just my theory how this will play out in the long run, I'm no oracle by any means.
  - CharlieDigital 9mo ago
    > that offer the same for 10x less money. Not likely because there is still a lower bound. These 1 person startups are winning partially because they are already 10x cheaper than the incumbents. But beyond that, it's not likely that there are 10x the number of people who know the domain and have the right mindset plus appetite for risk.
    - vidarh 9mo ago
      I'm not entirely disagreeing. There are limits there that means we can't assume the margins will go to their theoretical minimum. But you're also in part assuming the models don't increasingly know the domain or know how to research the domain and compile the information for you. They'll be squeezing margins out of a lot more than just the tech.
  - skeeter2020 9mo ago
    Discounting Apple, their products and their customers to a cult is at best jealousy but still blatantly wrong. Lots of competition has been trying to out-Apple them for decades with no luck, and it's not because an iPhone customer is stupid & brain-washed.
mikkupikku 9mo ago
I predict that the commercial market for a lot of software will evaporate as people find that getting AI to whip up a custom solution that fits their unique problem space like a glove is actually cheaper and simpler than trying to make COTS software do the job. We're not quite here yet, but maybe in a few years.
- DoctorOW 9mo ago
  > I predict that the commercial market for a lot of software will evaporate Counterpoint: Windows, Oracle DB, etc. have had free/cheaper alternatives for decades and still thrived.
- BikiniPrince 9mo ago
  Yes/no. Regardless of the code complexity reduction there is still architecture, planning and implementation. Could someone come by and clone my work afterwards? Absolutely. Will they retain customers with only a little understanding of the product or model? Questionable.
- 15155 9mo ago
  You aren't just buying software, you're offloading liability of continued support and functionality.
  - mikkupikku 9mo ago
    Sure, but there's a whole lot of businesses already using custom solutions made with excel/access/etc that are held together with duct tape and chicken wire, so I think the adventurous spirit necessary is there.
    - pc86 9mo ago
      There have always been hundreds or thousands of companies that want software engineers but simply don't have the revenue to support them. My first dev job was a small private company in exactly this spot. They basically paid me my salary for six months to figure out WordPress and PHP on the job having only ever done very basic programming stuff on my own in high school ~6 years prior. The median dev salary across the entire US is something like $130k/yr. There are huge numbers of new or self-taught software devs in low cost of living areas of the country making $50-60k/yr.
vidarh 9mo ago
I think so too. But in the meantime there is a quiet goldrush for people who spot niches where they can extract decent (or a lot) of value right now, and for long enough to be worthwhile. If they can get scale enough that thinner margins makes for a worthwhile business when the market catches up, great. If they can't, then we stay lean we might make off with decent ROI. But that is also a reason to be cautious of chasing capital and think hard about whether you can spend it sensibly fast enough to improve your own ROI... E.g. I have a project right now where I won't consider taking VC cash because I don't think I can spend it fast enough to buy me enough additional leverage to make enough additional money to compensate for the dilution and the other usual shenanigans before I expect margins will be squeezed out of the niche in question. It also means I don't think the opportunity will ever scale above a certain level, but that's fine - it'll be a quick attempt at grabbing what profit I can. Also, while we of course shouldn't diminish the potential moat created by understanding the product in favour of only value the tech, we need to also consider that AI's are a levelling factor there too. Claude knows (I've verified what it's said) more about the niche I'm vaguely talking about than I do - it knows pricing, it knows positioning/marketing, it knows conventions and requirements of the niche, and while I'm sure I could have found all of it myself starting from scratch too it shortcircuited an enormous amount of effort to get an infodump that let me know precisely what to look for to verify it. A lot of tech companies will find the institutional knowledge they thought would shore up their moat is worth a lot less than they thought.
- pkreg01 9mo ago
  > A lot of tech companies will find the institutional knowledge they thought would shore up their moat is worth a lot less than they thought. I totally agree. I think going forward the primary value of SAS will be the embedded domain expertise in a pre-built product. The comparison of Asana versus Notion comes to mind for project management. Asana forces abstractions of good project management upon you, whereas Notion lets you build it yourself. I think this principle will scale to all software in the future, where the only real value of software or it becomes exported maintenance obligations and a predetermines domain abstraction. But as you mentioned, I think companies will rapidly find that their own specific abstraction is worth a lot less than they believed.
CuriouslyC 9mo ago
This is mostly correct IMO. SaaS is extremely vulnerable, companies will be able to modify open source tools to do exactly what they need, and agents will make managing those services easier. This will lead to downward pressure on SaaS prices, and cause them to become more like cloud data management platforms that they let customers build on top of rather than one-size-fits-all apps.
- pkreg01 9mo ago
  I agree with this completely. I forsee an era of enterprise level 'template' saas products that are expected to be tinkered with and highly customized. I think products like Notion that have an incredibly robust customizability and integration layer are going to thrive, where every single company can use a template engine to build extremely customized applications - and the barrier to building on top of these will essentially become the rate of human speech.
lab14 9mo ago
Perhaps for extremely basic products. Most non-engineers can barely write and untangle their messy thoughts and you think they can just build a spec for an AI to build a product? Hopefully I'm wrong, but I doubt it.
- zdragnar 9mo ago
  This is what gets me... Even at companies with relatively small engineering teams compared to company size, actually getting coherent requirements and buy-in from every stakeholder on a single direction was enough work that we didn't really struggle with getting things done. Sure, there was some lead, but not nearly enough to 2x the team's productivity, let alone 10x. Even when presented with something, there was still lead time turning that into something actually actionable as edge cases were sussed out.

CharlieDigital 9mo ago

You are discounting sales, marketing, and branding. Take drop shipping for example: anyone can do this, but the successful ones are those that know how to brand and market the product well. Not to mention having the right mindset for startups and building a business. The code and product is maybe only 20% of the story.

margorczynski 9mo ago
I'm not. That edge eventually converges to 0 when you have 10+ competitors that offer the same for 10x less money. If you don't have some kind of cult following like Apple eventually you'll get margin-squeezed till death and all that marketing, sales, etc. will get cut down to stay afloat. Of course all of the above is just my theory how this will play out in the long run, I'm no oracle by any means.
- CharlieDigital 9mo ago
  > that offer the same for 10x less money. Not likely because there is still a lower bound. These 1 person startups are winning partially because they are already 10x cheaper than the incumbents. But beyond that, it's not likely that there are 10x the number of people who know the domain and have the right mindset plus appetite for risk.
  - vidarh 9mo ago
    I'm not entirely disagreeing. There are limits there that means we can't assume the margins will go to their theoretical minimum. But you're also in part assuming the models don't increasingly know the domain or know how to research the domain and compile the information for you. They'll be squeezing margins out of a lot more than just the tech.
- skeeter2020 9mo ago
  Discounting Apple, their products and their customers to a cult is at best jealousy but still blatantly wrong. Lots of competition has been trying to out-Apple them for decades with no luck, and it's not because an iPhone customer is stupid & brain-washed.

mikkupikku 9mo ago

I predict that the commercial market for a lot of software will evaporate as people find that getting AI to whip up a custom solution that fits their unique problem space like a glove is actually cheaper and simpler than trying to make COTS software do the job. We're not quite here yet, but maybe in a few years.

DoctorOW 9mo ago
> I predict that the commercial market for a lot of software will evaporate Counterpoint: Windows, Oracle DB, etc. have had free/cheaper alternatives for decades and still thrived.
BikiniPrince 9mo ago
Yes/no. Regardless of the code complexity reduction there is still architecture, planning and implementation. Could someone come by and clone my work afterwards? Absolutely. Will they retain customers with only a little understanding of the product or model? Questionable.
15155 9mo ago
You aren't just buying software, you're offloading liability of continued support and functionality.
- mikkupikku 9mo ago
  Sure, but there's a whole lot of businesses already using custom solutions made with excel/access/etc that are held together with duct tape and chicken wire, so I think the adventurous spirit necessary is there.
  - pc86 9mo ago
    There have always been hundreds or thousands of companies that want software engineers but simply don't have the revenue to support them. My first dev job was a small private company in exactly this spot. They basically paid me my salary for six months to figure out WordPress and PHP on the job having only ever done very basic programming stuff on my own in high school ~6 years prior. The median dev salary across the entire US is something like $130k/yr. There are huge numbers of new or self-taught software devs in low cost of living areas of the country making $50-60k/yr.

vidarh 9mo ago

I think so too. But in the meantime there is a quiet goldrush for people who spot niches where they can extract decent (or a lot) of value right now, and for long enough to be worthwhile. If they can get scale enough that thinner margins makes for a worthwhile business when the market catches up, great. If they can't, then we stay lean we might make off with decent ROI. But that is also a reason to be cautious of chasing capital and think hard about whether you can spend it sensibly fast enough to improve your own ROI... E.g. I have a project right now where I won't consider taking VC cash because I don't think I can spend it fast enough to buy me enough additional leverage to make enough additional money to compensate for the dilution and the other usual shenanigans before I expect margins will be squeezed out of the niche in question. It also means I don't think the opportunity will ever scale above a certain level, but that's fine - it'll be a quick attempt at grabbing what profit I can. Also, while we of course shouldn't diminish the potential moat created by understanding the product in favour of only value the tech, we need to also consider that AI's are a levelling factor there too. Claude knows (I've verified what it's said) more about the niche I'm vaguely talking about than I do - it knows pricing, it knows positioning/marketing, it knows conventions and requirements of the niche, and while I'm sure I could have found all of it myself starting from scratch too it shortcircuited an enormous amount of effort to get an infodump that let me know precisely what to look for to verify it. A lot of tech companies will find the institutional knowledge they thought would shore up their moat is worth a lot less than they thought.

pkreg01 9mo ago
> A lot of tech companies will find the institutional knowledge they thought would shore up their moat is worth a lot less than they thought. I totally agree. I think going forward the primary value of SAS will be the embedded domain expertise in a pre-built product. The comparison of Asana versus Notion comes to mind for project management. Asana forces abstractions of good project management upon you, whereas Notion lets you build it yourself. I think this principle will scale to all software in the future, where the only real value of software or it becomes exported maintenance obligations and a predetermines domain abstraction. But as you mentioned, I think companies will rapidly find that their own specific abstraction is worth a lot less than they believed.

CuriouslyC 9mo ago

This is mostly correct IMO. SaaS is extremely vulnerable, companies will be able to modify open source tools to do exactly what they need, and agents will make managing those services easier. This will lead to downward pressure on SaaS prices, and cause them to become more like cloud data management platforms that they let customers build on top of rather than one-size-fits-all apps.

pkreg01 9mo ago
I agree with this completely. I forsee an era of enterprise level 'template' saas products that are expected to be tinkered with and highly customized. I think products like Notion that have an incredibly robust customizability and integration layer are going to thrive, where every single company can use a template engine to build extremely customized applications - and the barrier to building on top of these will essentially become the rate of human speech.

lab14 9mo ago

Perhaps for extremely basic products. Most non-engineers can barely write and untangle their messy thoughts and you think they can just build a spec for an AI to build a product? Hopefully I'm wrong, but I doubt it.

zdragnar 9mo ago
This is what gets me... Even at companies with relatively small engineering teams compared to company size, actually getting coherent requirements and buy-in from every stakeholder on a single direction was enough work that we didn't really struggle with getting things done. Sure, there was some lead, but not nearly enough to 2x the team's productivity, let alone 10x. Even when presented with something, there was still lead time turning that into something actually actionable as edge cases were sussed out.