which industry do you think is the riskiest to start a business, and which is flourishing?
With AI, it seems to be everything possible:
Everybody is a builder, maker, marketer, founder, (vibe)coder/developer.
We are experiencing the "surplus" of AI products... it is harder to differentiate (our time, resources, and attention are limited).
Is there any industry that you consider to be disadvantageous for business in the current situation, and vice versa, what would be worth getting into?
Maybe I just took it from the wrong end and see things too pessimistically 😀, but I would still like to know your perspective on the situation and maybe see any empty gaps that can be filled.
What I've noticed, for example, is that things related to defense are doing extremely well now, also because of the geopolitical situation in the world.
For example, an Estonian VC firm Darkstar invested in purely military defense tech, especially from Ukrainian startups. The fund recently closed its first €15M round (target: €25M) and backs startups like FarSight Vision (geospatial drone tools) and Deftak (drone ammunition).
Replies
Services, definitely.
Especially in smaller IT service companies that are often founder-dependent. A lot of friends have mentioned that sales are drying up, partly because of the AI hype. There’s this growing (and false) perception that low-code or “vibe coding” tools can replace full-fledged enterprise solutions. That’s making it riskier to sell actual software services right now.
minimalist phone: creating folders
@chaosandcoffee Do you mean that the service is the riskiest? I would say there will be more products, but I also have in my mind that the majority of people are "lazy" to do something, so they will delegate it to people who are willing to do it.
Business in non-specialized SaaS is risky due to market saturation and security concerns, making it difficult to stand out or scale safely.
On the flip side, Generative AI is flourishing, driven by enterprise adoption and growing demand for real-time, secure solutions. Entrepreneurs and consultants tapping into these spaces will find opportunities for growth and innovation.
minimalist phone: creating folders
@vivek_sharma_25 But isn't there too much AI? I think once someone comes up with gen AI, it will be easier for everyone to generate anything.
“If AI can build it in a weekend, why should I care?”
If something’s that easy to spin up, it’s probably not defensible. Novelty isn’t value.
Example: yet another AI summariser? Dead on arrival.
But a tool that plugs into a hospital’s EHR system and translates medical notes into structured data — that’s messy, valuable, and hard to replicate.
The real plays now aren’t cheap clones, they’re bets on infrastructure, integration, or identity. Products that are hard to copy because they’re tied to real distribution, real trust, or real-world complexity.
minimalist phone: creating folders
@adam_martelletti Things that cannot be replicated so fast and easily will dominate the market if they are useful.
For sure brick and mortar businesses - they're usually behind the times on tech / automation, and could see a significant benefit. They're also not very tech savvy, so it could be a harder sell - having to explain how it works - but also potentially a higher willingness to pay instead of DIYing it.
minimalist phone: creating folders
@wesley_liaw yeah, many people do not want to work physically. But China is quite good at creating robots that can be pretty skillful (and sometimes dangerous) – remember the video when one decided to attack the audience :D
@busmark_w_nika Terminators are on the way haha! Sooner or later..
Great question. I’d say influencer marketing is risky yet flourishing at the same time. So many pain points still exist—ghosting, hidden pricing, and slow negotiations. That’s why we’re working on an auction-based model for sponsorships, hoping to reduce some of those risks for both brands and creators.
minimalist phone: creating folders
@uxyemi Do you mean BidVid?
@busmark_w_nika Yes, BidVid. We've just released beta to the public ahead of launch https://bidvid.io.
minimalist phone: creating folders
@uxyemi I hit the "Notify me" button :)
@busmark_w_nika Oh cool, thanks. See you at the launch 🙂
Honestly adaptive reuse properties like churches-to-mixed-use conversions are having a weird moment right now. The bones are usually incredible, we're talking solid construction from eras when they didn't cut corners, but the repositioning play is tricky as hell. I stumbled across a 20k sqft former church in Harvey, IL (https://realmo.com/listing-m/14701-myrtle-ave-harvey-il-60426/2241338940115257163/) that's already set up with separate HVAC zones and multiple entrances, which is exactly the kind of infrastructure you need for multiple tenant types. The math on paper looks solid (daycare + office suites + event space), but the success really hinges on local demand and whether you can actually fill those spaces consistently. Real estate's always been about location and timing, but right now I'd argue specialty conversions sit somewhere between high-risk retail and safer multifamily plays.
minimalist phone: creating folders
@john_bishopp I semi-agree with this statement. I think that restaurants were the riskiest during COVID. Anybody who survived that one event will likely survive anything.
I think LLMs have done a good job in solving multilingual translation tasks, but they have not yet permeated more actual cross-language and cross-cultural communication scenarios. Human translation companies (language service providers) are dangerous, but the opposite side is worth getting into.
I totally get what you’re saying. It feels like every day there’s a new AI tool launching, and it’s getting harder to tell what’s actually useful. Everyone’s building something, so standing out is tough.
Right now, I think starting something in the general productivity or AI-helper space is pretty risky. There are just too many similar tools, and unless you have something super unique or a very specific audience, it’s easy to get lost.
On the other side, I agree that defense tech seems to be growing fast, especially with everything happening around the world. Also, industries like healthcare, legal, and security are still figuring out how to use AI in a safe and trustworthy way. That feels like a good space to explore.
Lately I’ve also noticed more tools that mix automation with human input. Not fully AI, but AI that helps you work faster without taking over everything. Those seem promising too.
You’re not being too negative at all. These are real challenges. It’s just about finding where there is still real need, not just hype. Curious what others think too.