This question reappears often on HN. Note that you should submit as an "AskHN:" for responses to a question.
At any rate, here is my somewhat promising (but not guaranteed) look at the future of subscription software services--and I did not use an LLM to inform my answer:
My crystal ball says...
1. For software products that can be recreated in a week or less using an LLM and useful for individuals or small teams, then yes the future looks bleak. Examples include:
- to-do lists
- CRM
- anything with a simple database or spreadsheet backend
- messaging and comms
- conversions
- routine image processing or generation
- dashboards
- simple control systems
- simple, repetitive operations
- ...and so on
Why? Because these types of apps are easier to manage and support; and let's face it, most of these have had open source alternatives for years.
2. For more complex software applications--or software tied to hardware systems--the story may be a little different. Anywhere proprietary algorithms, techniques, processes, systems, etc. are involved, it will be a story of AI-aided development and support with some human involvement. Thus, subscriptions have a larger moat in this area.
3. Human-Software Services: wherever a software subscription is tied to a human relationship, there will be a greater moat for SaaS. These types of apps are uncommon now, but expect to see a significant emergence of "software+service" apps in the near future. Examples might include:
- TurboTax plus an accountant
- Canva or Adobe plus a designer
- Automobile service plus software
- Concierge service plus software
- Travel service plus software
- ...and so on
Why? AI makes customer service from a human easier, and human-to-human connections are more durable, experiential and satisfying. This higher value will support the SaaS subscriptions, and people may be willing to pay even more for the higher value human service.
AI will not replace humans--we can already see the limitations regardless of the amazing capabilities. But it will change the economy and drive more AI-supported, service-related work. It will take a few years for this transition to play out.
This question reappears often on HN. Note that you should submit as an "AskHN:" for responses to a question.
At any rate, here is my somewhat promising (but not guaranteed) look at the future of subscription software services--and I did not use an LLM to inform my answer:
My crystal ball says...
1. For software products that can be recreated in a week or less using an LLM and useful for individuals or small teams, then yes the future looks bleak. Examples include:
- to-do lists - CRM - anything with a simple database or spreadsheet backend - messaging and comms - conversions - routine image processing or generation - dashboards - simple control systems - simple, repetitive operations - ...and so on
Why? Because these types of apps are easier to manage and support; and let's face it, most of these have had open source alternatives for years.
2. For more complex software applications--or software tied to hardware systems--the story may be a little different. Anywhere proprietary algorithms, techniques, processes, systems, etc. are involved, it will be a story of AI-aided development and support with some human involvement. Thus, subscriptions have a larger moat in this area.
3. Human-Software Services: wherever a software subscription is tied to a human relationship, there will be a greater moat for SaaS. These types of apps are uncommon now, but expect to see a significant emergence of "software+service" apps in the near future. Examples might include:
- TurboTax plus an accountant - Canva or Adobe plus a designer - Automobile service plus software - Concierge service plus software - Travel service plus software - ...and so on
Why? AI makes customer service from a human easier, and human-to-human connections are more durable, experiential and satisfying. This higher value will support the SaaS subscriptions, and people may be willing to pay even more for the higher value human service.
AI will not replace humans--we can already see the limitations regardless of the amazing capabilities. But it will change the economy and drive more AI-supported, service-related work. It will take a few years for this transition to play out.
I will put my crystal ball away now.