Ask HN: What business would you start in 2025? | Hacker News Reader

Ask HN: What business would you start in 2025?

26 pointsjamesq49 comments

If you were to start from scratch, $0/£0 in the bank, what industry or business would you tackle? I’m just curious if AI would feature or if this is a saturated market.

View on HN

I would need people, so if I have no cash I'd start something holistic or FOSS and recruit volunteers that way. Eventually a community will build around it, and it will start to see some uptake. With some clever rewording of the license, I'd make the company public, pump up it's worth, and then cash out big before it collapses. Rince and repeat, sleeping soundly at night via various pills.

In 2025? Probably a consulting business to help companies replace their software engineers with AI.

In 2024 it would be too early, and in 2026 it would probably be too late.

LOL

2026: how to hire the right people to fix the mess of a code you have accumulated with AI.

this guy knows what hes doing, and he fcks too!

If I'm feeling conservative, I would open a self-storage facility.

If I'm feeling energetic and risky, I would evaluate the commercial viability of radioisotope batteries.

That’s a good shout - maybe storage business to find the radioisotope research! I keep hearing laundrettes are good conservative businesses too.

Ha, OMG, if I had a nickel everytime my partner mentions the idea of opening a laundrette. It's almost to the point of annoying... but color me intrigued:)

I can help with the latter. The math can be quite sobering when you look at the potential regulatory burdens.

Recently I've been thinking it would be a fun and potentially successful challenge to start a non-alcoholic brewery. Problem is, I don't know hardly anything about brewing.

Specifically though, I have struggled with the fact that many "third places" in my city are breweries - there isn't anywhere to go if you want a large, well maintained outdoor space with seating areas, that also has food and drink options, but that isn't a restaurant looking to turn over tables.

A good NA brewing space could provide the casual "come hang out with your friends/kids" third place that a good brewery does, while catering to those looking to reduce alcohol intake or those that don't drink.

This is actually a decent idea. I have a friend that’s in the brewing space running a small local brewery and can’t do non alcoholic as it is too capital intensive at his scale. But his customers want options.

Meaning there is a decent moat and customer demand.

Something in the mental health space to tackle the loneliness epidemic.

Focus on digital nomads.

Again and again they post and comment about loneliness, for obvious reasons given the lifestyle.

Why do they keep living that way then in your opinion? Being a digital nomad is something only people in very specific conditions can even afford to do so it’s a deliberate choice.

It’s hard for me to have much sympathy for someone who’s deliberately living a life that’s making them feel lonely.

Several reasons:

1. They're sold on a lie. Here's a photo of my laptop in front of a beautiful pool, next to a delicous bowl of $1 Pad Thai. Ignore the February burning season and total inability to communicate with anybody outside of self-deceiving 20-something Westerners like myself.

2. There are benefits. Being a DN is freeing and fun, especially if this is your first shot at remote work. You don't justify to anybody what time you have your sandwich in the morning.

3. Most DNs have never experienced actual loneliness before, so they don't know how crippling it is. They think all their friends and family having been away/busy one weekend or even one month was loneliness. That's like how the people in 5-day COVID hotel lockdown thought of themselves as having experienced imprisonment. Sure, only in a glancing sense.

4. Sex/romance tourism. Look at the hot exotic naive foreigner who's here to exchange some emotions/experiences with you, the hot exotic native person who's otherwise boring or too slutty/free-thinking for people in your own country.

Being a Digital Nomad is a life experience that comes with as many negatives as positives. But there are positives.

Why do people do it? Because the alternative is to never have done it.

I’d really like to see something like Meetup but… better. Not really sure how. Like, I don’t know, you pick a date/time window and get paired up with five random people to go and debate what the best cuisine is at a bar. An app just to create low stakes, spontaneous and silly interactions for you. I haven’t used Meetup for a while but I always found it so dry, and it was rare anybody from my age group attended, probably because the meet-ups are so boring.

In fact I just checked Meetup and all I see is weird scam courses and speed dating events.

Hell I’d love something like that even for random online interactions. Getting paired with random interesting people from all over and have casual exchanges that are private and not recorded forever on some social platform.

I'd pick a SaaS category with plenty of existing (non-VC) incumbents, and build something better that I personally have experience with using, just as I did in 2021.

Why specifically non-VC?

I’m working on something which I think would be a great product but a direct competitor got $25 million in VC funding in 2021.

With all the hype around AI, deeptech, etc., I’m wondering if there’s a market for R&D consulting as a form of technical due diligence. I see money being thrown at some of the most hare-brained nonsense that could have been avoided with a bit of critical analysis. I would think sophisticated investors and customers would find this kind of service useful, but maybe I’m missing something.

Does anyone have experience with this kind of work?

There is a market for technical due diligence consulting, but the work is typically done by friends of investors; it's a difficult and tiny market to get into.

Fresh bread delivery. There’s nothing worse than waking up and find out that you dont have bread. Would be great to have fresh bread at my door every morning.

Something blue-collar-ish, like cleaning, plumbing, construction, repairs. Anything 'trade'. People are significantly less self-sufficient nowadays, and setting up a plumbing business (in my country, The Netherlands) would easily get me a million-euro business without too much effort.

If it has to be in tech, I would jump on the latest trend and try to squeeze it quickly before moving on.

Let me by your Poolse schilder haha

Making systolic array chips (my BitGrid design) to bring provably secure Petaflops to the masses. They're handy for any algorithm that can tolerate extreme pipelining, including deep nets.

Because everything is trivial to reason about, there's nowhere for bugs, or zero day exploits, to hide.

Oh.. and we'd do an open source data diode product as well.

Do you have any other examples where systolic arrays are suitable other than matrix multiplication? As far as I am aware, other problems require different systolic architectures. So I am curious whether you are talking about a general purpose architecture.

I'm talking about a Turing complete, general purpose architecture.[1] One with some seemingly stupid choices that turn out to work well. A cartesian grid of look up tables (LUTs) with latches, clocked in alternating phases. This slows things down, and makes it completely deterministic, and very easy to reason about. There are no race conditions to worry about. Like an excel spreadsheet, it's very easy to see all of the dependencies, and circular references aren't possible in the traditional sense, because of that delay.

I'm stuck in analysis paralysis... or I'd have more than an emulator[2] to show you. In theory, you can take any expression that can be broken down to a directed graph of binary logical expressions, and compile it into the "program" for the BitGrid. Because the grid is homogeneous, you can shift or rotate, or flip it to move the I/O around. The aforementioned dependency tracing makes it possible to prove the functionality conforms to the desired logical expression graph.

I need a kick in the pants to get the rest of this thing figured out. As near as I can tell, the only problems are my own writers block.

[1] https://esolangs.org/wiki/Bitgrid

[2] https://github.com/mikewarot/Bitgrid

FOSS patent troll company. Agents would get employed by random companies that use linux/oss. Scan the licenses company uses, report back to HQ. First email would be from marketing "we can fix your issues", in case no or negative response, email from lawyers would follow.

Non-genAI/LLM applications of ML are most exciting to me at the moment. Seems like the tooling and techniques are at a good point of proven but with creative prospects not yet explored, especially with generative siphoning off some of the hype. Worst case I’d learn some stats.

People sleeping on the fact that Deep NNs can approximate any function. Let them focus on hallucinating chatbots, make something actually useful!

The economy is bad, everything is expensive, and there is absolutely no social climate for emigrating within EU. That would be scenario equal to having my country invaded, or getting cancer, so death rather than thinking about opening a business. AI is only a topic in huge enterprises, I have yet to see anyone starting from zero and earning any money with it. There is lots of money incoming to real estate, physical gold, football, and freak show fights, so if you don't mind being predatory asshole and dealing with organized crime there is potential money stream to tap into.

I think economic decline is regularly considered the best time to start a new business?

Good time to find people to hire and work hard to prepare for better times.

wait until your first full bankrupt failure before confirming; extra credit, add a financial disaster divorce

Yeah easy to say of course. I am sorry for your troubles.

Isn't this a whole concept of a crisis? You expected to buy half of a country and hire another half, but instead you're among those who are struggling.

If everything is expensive, can it be possible to start a new company that will make some things cheaper?

We built the world on a foundation of crates of dynamite. We're now in the process of demanding better packages and purer ingredients. We want the users to be more careful.

I'd suggest selling bricks. In other words, operating systems that never, ever, trust user code. See Genode or GNU Hurd for examples.

I have a feeling I would do something non-tech, I have a few ideas for FMCG items but I am not sure how to produce those at scale cause well its not my area of expertise, nor do I know about supply chain stuff. In terms of working capital, I don't have enough to produce it in meaningful numbers.

"FMCG stands for Fast-Moving Consumer Goods products sold quickly and relatively cheaply. Some examples of FMCG products include packaged foods, beverages, toiletries, cosmetics, cleaning supplies, and other low-cost household items."

Security consulting for AI use... The boring traditional stuff as main focus. You know making sure that dialog box that you input your prompt is correct or you handle user accounts and so on correctly.

On side could also sell testing against AI, but the usual suspects are already enough work.

Ask yourself `what does everybody want/need? (Hint it's not AI) Two (of many) possibilities; accommodation & food. One could investigate previous examples where startups began small and made history. Then I suggest you may need at least 2 of these: #1 A gimmick #2 Good luck #3 $$$$$$.

I agree with your 2 of 3 suggestions…

There's plenty of people doing accommodation and food. Or anything else with high demand. The profit margins are tiny because of this, aka red ocean. More of these don't always make the world a better place; you just end up fighting with a smaller competitor, while Big Corn rakes in the profits from the economy of scale.

and add the 3rd answer as well: everyone wants money too. Make something to help make people money. Trading alerts, charting tools, stock market aggregators, crypto aggregators, anything that can push people in that direction

Palliative care and interstellar journeys.

Honestly, my idea would be a privacy protecting search engine, think DuckDuckGo, that promises to never use AI without careful research and if it is used promise that it'll always default to off.

It feels like everyone is putting AI in even though not many people particularly want AI.

You put AI in the description to grift money out of investors hoping to grift other investors, same as blockchain a few years ago.

Using the OpenAI ChatGPT APIs doesn't make you an "AI" company, but they don't care.

A No Code platform. Yeah, I know it is a crowded space. There aren't that many great options yet tho, enterprise NoCode platforms are all bad, they don't know how to make good software.

Patching together a bunch of random React, NextJS, Prisma code won't do it for much longer.

I dont have much information.In my humble opinion,a nocode platform based on python and excel or something would be easy for AI tools to ccreate tools since they are quit efficient in python and excel coding . P.S :Info might be a bit outdated

Yes, it is easy to create low quality things, quite hard to create anything professional

Reply on news.ycombinator.com