Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Is finding gaps just hard now? As a 24 yo, everything seems to be done already
26 points by wxce on May 9, 2024 | hide | past | favorite | 47 comments
I'm trying to find good problem statements and gaps to solve, Tech/Non-Tech.

But it just seems to hard find something substantial. Everything seems to be already done.

The gaps seem very teeny-tiny. I really don't want to solve a hyper-specific, niche B2B sales gap for example.

It seems like people earlier had it much easier and the ideas seemed a bit obvious. This sounds envious, and could be somewhat wrong. But I truly am still trying to figure out a good gap.




Startup philosophies become outdated fast.

Early 2000s it was: just put it online and it will sell, at a time when online businesses were still new.

Paul Graham's Y Combinator was: create a good product and it will sell, at a time when online businesses were these half-assed 2000s era products.

AirBnB/Uber's generation is: get a fuck ton of funding and do mildly illegal shit fast, and then use your funding and lawyers to rewrite the law. So called "disruption".

(I am not going to mention OpenAI and other AI startups because they are a class of their own.)


There is a whole SHITLOAD of inefficiencies in the way companies are run. You have to be tapped in and talk to a bunch of people and network to see this, but once you do, and you have technical knowhow to solve these problems, you can easily make 1 mil a year selling these solutions to companies because in turn it saves them money. And most of them don't even require AI.

There is a reason why Zapier has 5 billion dollar valuation, and its basically just a bunch of api calls from a technical perspective.


> There is a whole SHITLOAD of inefficiencies in the way companies are run

My jaded and cynical ass assumes those inefficiencies are deliberate so that middle managers can justify their jobs.

Everyone claims to hate bureaucracies, but there are some people that are profiting from maintaining them.


I know a lady whose entire job was literally replaced by an Excel spreadsheet that I created in a few nights of intense VB programming. It wasn't my intention to "eliminate jobs". Anyways, I got basically nothing "extra" from the company for that extra-credit work. Moral of the story is that there are plenty of problems to be solved but our compensation is only based on what we can negotiate.


Most of the time, the only thing we get for hard work is more work. This is not to say that we shouldn't improve things in our jobs or remove inefficiencies - this is just to say each individual should learn to draw the line somewhere and should be comfortable with that line.

"Hey I did extra at my job, but nobody cared. That's okay, no big deal" <-- this is good for one's mental state.

"Hey I did extra work at my job and improved process/saved money. Nobody cared, this sucks, my career is crap, my life sucks..." <-- time to either do less at the job or find a better job


Well there are still pen and paper companies doing fax. There are some old timers running not so small companies just because they don’t know better.


There are gaps everywhere, especially in science/research.

Consider looking into a Ph.D. if you’re hungry to find, master, and close a gap. Not being sarcastic, btw.

The more you know, the more you realize how little you know and how many gaps there are in your field.


No. Absolutely No. See the chart labelled "Not So Very Serious Stuff" at https://philip.greenspun.com/careers/

Avoid the Ph.D., take a BS/BA and an MS if you have the time, but avoid even trying for a Ph.D.


? Where are the data from for his ascii graph? Your claim is absolutely contrary to everything I’ve read.

Opposite sources: https://grad.msu.edu/phdcareers/career-support/phdsalaries https://www.wes.org/advisor-blog/salary-difference-masters-p...


Wow. I haven't even thought about Philip in almost 20 years :-)


Do not do this. This is a trap. Research and higher academics are in effect a scam in modern day. At one point they were useful, but now everything done is overwhelmingly private R&D. In academia you will scrounge for scraps and never pay off the loans.


FWIW, I did my PhD in neuroscience and am now working in R&D in biopharma. It would be very difficult to have landed my current position without a PhD. Lastly, most STEM PhD programs pay a stipend (a small one, mine was 30k/year), so you’re not going into debt to get one.

I agree with your sentiment. Academia as a career path isn’t viable (or even feasible) for the majority of people who get a PhD


But most people doing research at private companies have a PhD at the very least, if not experience in academia, do they not?


Brain research

I watched a lecture series on the brain and my take away was that we really don't know "how the brain works"

Surprised me how much is unknown compared to say fields like physics/chemistry


Seconded: I am doing a PhD in the area that I previously created a start-up in, and I have a handful of small start-up involvements as well.


It’s hard to find problems without hands on experience in a problem space.

Then what you’ll see is there are A LOT of snake oil salesman (just pour some AI on it). There is also just a lot of bad quality/ limitations in almost all products. Solving practical, sometimes boring, problems, in a way that actually consistently delights users, is actually by far the exception.

Only by learning the domain, and trying to solve it, ideally with yourself as the main user, can you cut through all the BS and noise.

In other words care more about a domain first, be of service to people in that domain, try to solve problems with existing tools. You’ll see there gaps. The real problems/products will fall out naturally.


That and also having connections in business area.

From what I have seen in startup I worked for we most likely did not get a customer that my boss did not work with earlier.

Other ones were most likely choosing competition based on who they knew and worked with earlier.

Mind you we still had to go through due diligence before they signed, but getting foot in the door by knowing someone who can propose us as a vendor was big help.


> The gaps seem very teeny-tiny

That's how all markets start. Every product or startup needs to start off as a very targeted application because otherwise you will drown in complexity and lack of strategy, but with a forward facing vision on how to expand into adjacent problem spaces.

$TECHNOLOGICAL_APPLICATION for $SPECIFIC_INDUSTRY is always the best method to start building out a product, and expanding from there.

> Is finding gaps just hard now

No.

You might just not be looking hard enough. Try and find specific problems in a domain you are very experienced in.

Also be open to pivot if needed. For example, Slack started off as a chat-app within a failed videogame.


Here is a resource for you to examine:

"The Encyclopedia of World Problems"

https://uia.org/encyclopedia#:~:text=The%20Encyclopedia%20of....


Perhaps the low hanging fruits in this era of technology/software landscape have been poached. However, I can find hundreds of problems just in the realm of software architecture and infrastructure management that need to be addressed and improved.

Spend a decade in any industry and you will see the gaps more clearly.


Personal computers were a tiny market. Languages for them was a need, but it was only a tiny gap. Some guy named Bill Gates wrote a BASIC interpreter that was small enough to fit on one. It was a tiny niche... but it kind of grew.

Paul Graham points out that some niches are "the Microsoft Basic" of something much bigger. And, of course, some niches aren't. But if you're looking at is as "what thing that's bigger can this be the doorway to", you're more likely to find something that's more than just that little niche.


The number of software projects that have rough edges or missing integrations is astounding. Maybe those aren't glamorous things to work on, but it's definitely there.


Everything seems obvious in hindsight, and it takes a lot of work to make something that seems obvious and inevitable at the end of the day.

Think of how often you hear people complain about something. Every single complaint is an opportunity… something that hasn’t been solved well enough yet. The more people who complain, the bigger the potential opportunity.


OK - a big problem for starters - how to get people off of their screens and focus on something else - like life in general?

Given the way digital infrastructure exists as is today, this is a freakishly difficult problem.

In other words - how'd you make digital systems all around you "attention-enhancing" rather than "attention-diluting"


Isn't life in general effectively a solved problem by now? What more focus does it need?


How come? Basics such as lack of nourishment is not a solved problem in many parts of the world. And of course thats the bare minimum of material existence. Human potential offers way more, right?


Everything sounds obvious in hindsight. If you're one person, a small niche is probably where you should focus.

Here's my back-burner side-project: software to track materials being checked in or out, like a tool crib or maybe a library. Exists in many forms already, but mine's better :-)


There are pretty massive gaps.

Hiring people is really really hard (when was the last time you hired an expert to do something)? Coordinating people is really really hard. (These are problems facing every company, government, and group of people in the world). There is now way we're at a optimum.

Most systems were created, and then became stuck robots should - buy my groceries, move my money, book my events, schedule my events, suggest events, tell me what to buy, book my flights .... All of this is possible, it's just not automated :(. And frequently the barrier is data access.

More of my day is consumed by things that could be automated, but are not.


Sometimes more experience (in general, and in particular areas) reveals more subtle but important gaps, both in markets and in research.


education

is there anything that improves people's lives in the long-term as much as fixing education? there are plenty of niche sectors to excel on, feels like taking a candy from a baby

angels/VCs tell me it's too big of a market, thus competitive, thus you'll get eaten. I disagree with the last part -- we all say execution is key, but few of us realize what that means: the whole collection of decisions a company makes defines its 'execution', so each company is unique, even if they're redoing something that already exists. you can specialize so much no one will care competing with you early

as long as the inefficiencies in edu everywhere scream for help and the access to edu/work opportunity is not evenly distributed world-wide, there's huge room for innovation, and I'd add it's even a mandatory mission for us crazy visionaries to fix that


Kinda weird to ask when we're at the dawn of AI. There's relatively low hanging fruit everywhere right now thanks to LLMs, the industry will be digesting use cases of just the tech we have today for the next decade or more. It's not even that hard to get caught up with the research and start kicking out research ideas. Two weeks ago I was telling my father that the next step in transformer efficiency would be an algorithm that worked more like diffusion models, in which many tokens are predicted simultaneously with the model incrementally pushing them towards the correct answer. Just a day or two ago, CLLMs are announced and do exactly that. My maths isn't good enough to write that sort of paper but the idea was so low hanging that even a dabbler could work out what to do.

And that's ignoring all the bits of tech we use every day that could just be a lot better. JVM build systems are a mess, there is a serious need for competition there, to pick just one example that bothers me regularly.

Once you get into the swing of it the problem is not gonna be finding ideas. The problem will be finding ideas that other people will pay for. That's the hardest constraint.


Don't find gaps, simplify and modernize already available solutions. Few examples from the past. Telegram was created when Whatsup (and before that ICQ) was dominating the market. What's the point to create a message app when there are dozens already available?


At least in computing, the 70s-80s era was long gone. You either have to have a shitload of money or have a very good brain just to start.

It's indeed easier to be early than being smart, and of course those people are both.


> It seems like people earlier had it much easier and the ideas seemed a bit obvious.

In my experience: Everything always seems like it is already done, until something changes in the world, and then a bunch of obvious gaps emerge.

Which is to say that it is quite possible that earlier people did have it easier than what you are experiencing right now, but that doesn't mean it was always easy for them. They had to wait for the right moment.

The gaps will find you when they are ready. The bigger challenge is in being ready to accept them when they come. From what I have seen, when the obvious gaps start to come, you start getting pulled in a lot of different directions and, if not properly focused, can easily see the opportunity go by.


Things are always changing, new gaps forming. Is a lot of the low hanging fruit gone? Yea, but still a lot of opportunity out there.


Get a job at IBM/Deloitte or P&G and you'll see no end of things still left to do


Let's see: maybe try writing sentences that starts with "I wish I could".


To add to this, you must try to do something beyond being a software engineer to encounter more low hanging fruit.

Most of the easy “I wish I could” problems that are directly related to what a typical software engineer does on a day to day basis are indeed well solved, because that’s the lowest of the low hanging fruit.

Even going to mechanical engineering, chemical engineering, whatever, you’d immediately be like “wow there’s really no software to do X for you?” Not being an expert in the field will make it easy to underestimate the complexity of ideas, so be weary of that, but there are definitely still low hanging fruit ideas once you get off the beaten software engineer path.


Only slightly less lower: Make tools for unskilled labor. Go do the labor until you are familiar enough with it then talk with people who did it for decades. They found ways to make the simple things into impossible puzzles and solved them for the most part into a convenient step by step process. The software then assists the proverbial noob (without being annoying or getting in the way) and smoothens the learning curve.

How many ways are there to wash a car and when should you pick which?


https://www.fastcompany.com/3001329/how-move-past-everything...

Here's an article from 2012 on the subject of "everything seems to be done already"

A few years after this article was published. Google did "AlphaGo" and later "Alphazero" for chess.

OpenAI was founded in 2015.

OpenBCI kickstarter was in 2014. They are: https://www.youtube.com/watch?v=ZkWJem3LY5E

>I'm trying to find good problem statements and gaps to solve, Tech/Non-Tech.

What you're coming up against is starting to be better understood. Some people don't get ideas. Other people, that's all they get.

>It seems like people earlier had it much easier and the ideas seemed a bit obvious. This sounds envious, and could be somewhat wrong. But I truly am still trying to figure out a good gap.

Universities struggle with this a great deal. The people earning their degrees are often extremely underemployed because they cant come up with ideas OR they can't execute on the ideas.

I'm a fail to execute type of person lol. I have tons of ideas.

Fully automated agriculture building that takes water and electricity in and outputs food. Could be built anywhere, thus the extremes of droughty africa or arctic lattitudes.

Cultured meats, just the real deal type of meat, not this cancerous plant fake stuff.

Clean enough water production, less then 1% of the world's water is sort of clean enough.

Wearable tech is on the verge of finding something viable. Samsung/oura rings are kind of there, but what about a necklace?

Someone needs to find a good use for quantum computing.

Lots of room for next gen batteries, grid scale battery storage, or nuclear fusion work.

Plastic alternatives are in major need.

How about medical devices which never pierce into your body? hyposprays? sensors?

remote/wireless power distribution will be huge.

exoskeletons as mobolity scooters?

fully automated road construction that doesnt need employees?

a small drone driver/car with machine ai, that can drive around bike pathes and parks to clean up garbage? No people needed.

the electric airplane industry? That's going to happen in the next couple decades.

unmanned blimps that can move heavy cargo?

Who knows if there's any good ideas there. probably all trash.


>fully automated road construction that doesnt need employees?

I vaguely remember a solar road concept with tiles with a bunch of electronics in them. I suppose decent computers are cheap enough now. They could have some means of locomotion and drive fly or walk from the factory to their destination.

> unmanned blimps that can move heavy cargo.

nothing wrong with hydrogen. A tow boat with a long sausage behind it with containers hanging under it.


>I vaguely remember a solar road concept with tiles with a bunch of electronics in them. I suppose decent computers are cheap enough now. They could have some means of locomotion and drive fly or walk from the factory to their destination.

Solar freaking roadways! This was almost 10 years ago now. Feel old!


bicycle paths seems a lot easier.

They did build a road in China in 2017 but it's not modular it seems.

https://www.nsenergybusiness.com/features/china-solar-highwa...


if they are good or not depends on your perspective. Say, is good only profitable to you?

> Fully automated agriculture building that takes water and electricity in and outputs food.

i've pondered that one. it seems mostly a self- maintenance apparatus.

> clean enough water production, less then 1% of the world's water is sort of clean enough

what to do with the crap left behind?

> Plastic alternatives are in major need

we have bioplastcs they are just not popular. why i don't know.

> How about medical devices which never pierce into your body? hyposprays? sensors?

There is some cool tech that squeezes your legs to compliment heart beats. It apparently does miracles for health.

Rife technology works but no one wants to hear it. You get some of the most goofy debunking in the cosmos.

> exoskeletons as mobolity scooters?

seems a great idea.


>if they are good or not depends on your perspective. Say, is good only profitable to you?

Ive never even done bake sale. I have better ideas that I could likely make into a business but I have no idea how.

>what to do with the crap left behind?

Sell it? Bill gates open sourced a design for this. Easy enough to put together enough cash to build one. Unfortunately its just not viable.

>we have bioplastcs they are just not popular. why i don't know.

Same problem. I mean not-paper-straws. Like an alternative of plastics that is simply superior.

>There is some cool tech that squeezes your legs to compliment heart beats. It apparently does miracles for health.

I have experienced that for a brief time. No idea how good it was.


It’s not about gaps it’s about value


You are not wrong. It is a LOT harder today than it was 30 years ago. But....

30 years ago we had other challenges. For starters computers were "rare" (by today's standards). So you first had to convince buyers that "computerization" is a thing.

We also didn't have things like always-on networks, or OS-level printing. (Fun fact, my first payed gig was writing a program that could draw a diagram and print it on one specific printer model, an HP Laserjet 2.)

Anytime there is a new platform, (think dos/windows/web/mobile) there are opportunities to enter the market. For example we built a business app for windows circa 1997. It's been constantly developed since then. We gained a foothold in the early days, and are now a big player in our space.

But -entering- markets is harder. We entered a mature market with a different product in 2014. While we were a lot better, it was very (very) hard to break into the market dominated by 2 existing players. We have made progress, but frankly it's hard going.

So when looking for new things, it's helpful to first identify new markets. Timing this us hard, but if you hit the sweetspot it can work. That said, betting on a platform is risky (as Twitter API consumers discovered.)

You've already seen a flood of AI products, and you can expect a LOT more in the next couple years. It's the closest thing to a new platform in a while. Of course in 10 years that market will have consolidated and 99% will be gone. It's certainly a risky space, but worth a punt if it interests you. Hard to stand out though.

Finding new markets 30 years ago was easy. Finding new markets today is much, much harder. The niches are by implication much, much smaller.

That said niches exist. And getting a foothold in one niche can lead to it growing, or take you into another niche etc.

Selling into niches now is easy. My HP program was only useful for people doing genetics research, who submitted papers with genealogy diagrams, and happened to have an HP Laserjet. Talk about a niche market, Before those folk were on the internet...

Today I am, if anything, even more niche. I sell to a potential market of maybe 2000 people world-wide. I reach maybe 10% of that. But it makes a decent living. Not VC decent, but I do ok.

I would encourage you though by "trying things out". Find a niche, however small, and plug it. Then move in to another.

(Hint: at some point you'll discover that marketing matters more than coding. Getting good at marketing to niches is a killer-skill.)

Good luck.




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: