Most things are strange, missing, broken for mundane reasons that you can't do much about. The space of new ideas is huge, but almost all new ideas are bad. One strategy I really like is to think of technologies you couldn't build 5 years ago but are just barely possible today. That's where yesterday's bad ideas become today's good ideas.
For example, laptops and phones produced 20 years of incremental improvements in li-ion batteries, and suddenly electric cars are "unlocked".
"With a big enough engine even a brick will fly." - famous quote about the F-4, which captures that much of early airplane development was engine development rather than aeronautics. It's no coincidence that Glenn Curtiss got his start in motorcycles before moving onto aircraft.
I think to some extent we are seeing the same thing with the latest AI models, where if you just throw enough parameters and GPU's at a problem, you can get pretty good results.
In a way, I see some parallels to aircraft, in that it was important to study birds as the best example of heavier than air flyers early on, but understanding the nuances of the biomechanics of flapping wasn't necessary to surpass them.
> One strategy I really like is to think of technologies you couldn't build 5 years ago but are just barely possible today.
Streaming gaming is something I am still laser-focused on. Our networks are only getting faster. I feel there is a gigantic opportunity here that is being passed up by efforts like Stadia.
To do this thing correctly, the game engine itself needs to be developed with these concerns in mind. Expectations of things like having GPUs available on the same machine that is servicing player input events should be eliminated. The simulation itself may benefit from being distributed across multiple nodes connected via a low-latency network.
The future of shard-less, million player MMO worlds will not be feasible if every player's computer will be required to replicate and render all of that game state.
MMO seems to be in decline, not due to technical limitation. Also, I’ve just tried GeForce Now and I was blown away. I don’t think that huge changes in the model are practical in order to get 1% improvement over what’s available now
You can still do things like client-side prediction with streaming, but imagine if you controlled all of the clients and could always trust them. You could have the actual simulation running in a big bunker in Ohio somewhere on entirely synchronous terms (<1ms node-to-node), with a bunch of "player" nodes geo-distributed talking to the shared simulation and operating with CSP.
The game concept itself can be engineered around vertical scale constraints. For instance, imagine flight routes in WoW. If the cost to take a flight was modulated by the system load in the desired target region, you could create a much more seamless way to make this work. Think about real-world flight routes and system demand... The determinism goes down a lot for the technology owner, but the effect is much more compelling to the end customer.
I covered this in my PhD thesis, please check it out and I'd love to hear your thoughts as someone working on this (relatively lonely) problem! https://yousefamar.com/memo/notes/my/phd/
From what you've written so far, while the tech is cool, I'm not sure you're trying to solve a problem that actually exists. This is coming from someone who is bullish on both cloud gaming as well as decentralised architectures for gaming (esp VR) but for other reasons.
You seem to consider updating and rendering millions of players client-side to be a bottleneck; it's not -- human perception bottlenecks kick in long before that. Players can't -- and don't want to -- see all other players that are in the same area. And if they're not, then sharding by area is a much easier solution.
Even if it were a bottleneck, the solution is not to offload that rendering to cloud GPUs, it's many well-established LOD tricks. Game servers don't send millisecond updates of every player in a world to a client, the bandwidth they end up using for good UX is almost always lower than streaming video. You said networks keep getting faster, true, but client GPUs keep getting better and better too. Shard-less horizontal scaling sever-side has already been done btw (see SpatialOS (https://ims.improbable.io/products/spatialos) or WorldQL (https://github.com/WorldQL/mammoth)) or the many proprietary implementations (think Fortnite, Roblox, etc) but I don't think it's as useful as we think it is.
Besides this, I can tell you for a fact that "we have more players than we can handle" is an extremely rare thing actually, even on launch days, even though it might not seem like it to us. The biggest problem for games is getting players in the first place. The ones that do tend to have big budgets to spend on their servers, and it's usually not budgeting that's the bottleneck, but bad planning.
I would be curious to know if you've validated this idea with any studios, and if so, how? Or is the 10-20 hrs mainly focused on building?
Your thesis sounds very interesting. I'll give this a look.
> I'm not sure you're trying to solve a problem that actually exists.
This is more about creating the next generation of problems. I want to enable future experiences that HN would believe are comically-infeasible right now. The greater the disbelief the better.
I do have a very specific game concept and roadmap in mind. I've not talked to any studios or investors yet. I don't intend to until I've published the first game entirely on my own.
The hard part is all the art in a massive world. My current title is constrained to deal with this reality. I'll definitely need outside help on the real deal. Not wasting anyones time until I'm certain it's going to work.
True - you always have to deal with the final frontier. There are still very powerful statistical/hybrid systems that can be employed to mitigate cheaters.
You can combine many factors to close the last mile. Getting privileged information off the client machine is a huge part of the battle. The rest can be dealt with using clever tricks, stats, etc.
For example, imagine a game where you are using this ML aimbot to lock onto players heads. The developer could design a ramping detection system like:
1. Statistical detection - outlier in performance. Begin deeper analysis.
2. Review player inputs using our own ml models to determine likelihood.
3. Escalate to active measures - in-game canaries to bait the aimbot into very unlikely, inhuman responses.
Frankly I think the appetite for socializing online is dying.
If a machine in 5-6 years can leverage AI to render an entire unique universe of AI driven characters why sit in an MMO full of gold farmers, normalized, repetitive meme spammers aka other people who browse the same normalized social media?
Y’all have a good life, but I don’t need to know you all personally, nor did I ask you to solve these “problems” looking to be sold for coin.
> The future of shard-less, million player MMO worlds will not be feasible if every player's computer will be required to replicate and render all of that game state.
I think the problem here isn't render, on that scale you'll have to shard on the server end anyway.
A similar strategy is to find things which were invented 10–100 years ago but considered insignificant then, but due to changing environments can be adapted to be useful today in a way that wasn't possible then.
My day job is to work on software working with self-organising maps, which were last popular in 1980. As far as I know, they were never really used for anything significant then. The reason we can be so successful using them is that we discovered a relatively new environment in which they can improve profits fairly immediately: e-commerce.
The reason we prefer SOMs is that we have been able to get them to produce good results with less data than e.g. an SVD. They're also more open to domain expert diagnostics and tweaking.
'Ideas' are too early to be judged as good or bad.
Ideas are fodder, they are there to be moulded into things that are good or bad.
Often ideas are just foils to help you think.
They are so contextual as well: 'full screen smartphone' was maybe only possible by Jobs, with that much direct and coercive power over engineers. Motorola CEO was not going to pull it off. But Apple did work with motorola to make a 'music phone' as an experiment.
I think your note about 'yesterday's ideas' is bang on.
i think new ideas come from being naive and possessing the "no-mind" or a childlike sense of wonder. the older i get, the less plastic my brain is and the more depression and other issues start to seep in, get in the way, overcomplicate shit.
to be completely free, uninhibited, released from guilt and predjudice. to just think, play, laugh, fart, and yawn. boring days spent looking out the window wondering how crows fart, or what makes zebras sassy.
the only limit is our own adult psychosis. we are prisoners of our own minds, locked and shackled by constructs of our own design. too focused on the failures of the past and anxieties of the present to take a moment and b-r-e-a-t-h-e. like who fucking cares about businesses and startups and shit. saas my ass. when you actually stop trying, give up, and stop caring, and truly let go (like in the movie frozen) you can actually go in and through and back out the other side.
just learn to get out the way of yourself. that's what im working on, but it's really hard. you think you're supposed to be there and generate brilliance 24/7 but it's not possible.
go outside and find a patch of earth where things are growing. get really close to the ground and start looking at it real hard. there's honestly so much stuff there, it's wild. like who knew? wow thats nuts. acorns, rocks and shit. touch the dirt and rub it on your face. mushrooms are literally everywhere. the water in my body was once in a bowl of rice that your great grandma ate at becky's 27th birthday party.
I'm locked in my own personality trap. I don't understand why people buy shit and my toxic also think they are dumb sheep of some made up social values. And I can't change myself to get out of my trap. I always think what I'm going to sell to people has to be really good .. but I just realize it's about a lot of bullshit going on on marketing and sales and network effect vs pure functionality.
Here’s a quick tip: remove the word “bullshit” from your personality and assume that people aren’t morons.
People obviously do do dumb things, and some stuff is obviously crap. But if you change perspectives a lot of stuff actually….makes sense?
Like, looking at the world and seeing a bunch of bullshit is actually a flaw in your own thinking, not the world.
Take fast food for example. Widely derided, clearly poisons the body, not the best taste, produces a lot of waste. More expensive than cooking.
But, flip it around:
* Standard experience —> you know exactly what you’re getting
* Much cheaper than restaurant food
* Much faster than restaurant or home cooked food
* Tasty
* Easy
It’s something with obvious downsides but also obvious upsides.
It isn’t obviously dumb. It is only obviously dumb if you decide the downsides outweigh the upsides and then ignore all the upsides. Which is flawed thinking.
You can repeat this analysis and find that on their own terms most people mostly aren’t doing things that are totally irrational.
Even something like a lottery ticket gives someone hope despite the clear negative EV. You can take that too far, but….
The point is if you think everything is bullshit, your analysis is probably crap. And merely allowing yourself to use the term bullshit rewards this sort of lazy, negative analysis.
Whatever you sell doesn’t have to be the absolute best of all possible things. But it has to meet a need, and the vast bulk of stuff which sells does exactly that.
Appreciate your effort. But I would say food is not a really good example. I can relate taste via tongue, it's something very understandable. I don't think I can trust bias(reasons) after bought something.
> The point is if you think everything is bullshit, your analysis is probably crap.
Yes, it is crap. Because as I say I most values I can't relate .. a lot of why
> But it has to meet a need
That's the problem. Most of the time I don't think they need it.
Sure but that's an empathy problem. If you intend to sell anything to anyone you need to change your whole perspective.
For example, say someone goes buys a mac because they think the software is pretty and they like having things managed. You can't argue they are mistaken because you could get a linux machine with freedom for much less. They don't want such a thing.
Conversely, say someone buys a Linux machine because they like getting inexpensive tools and they enjoy fiddling with software. You can't say they are wrong because they could spend much less effort on computing by buying a mac.
There are no mistakes in either choice if they align with preferences. Here's an example of a mistake: buying a linux machine for ease of use, or buying a mac for maximal control over your machine.
So ask yourself: are the people making a mistake by having different preferences from you, or are they actually making a mistake on their own terms?
You don't have to cynically exploit your customers just because that's what other businesses do. If you can make an excellent product that solves a real problem and charge a reasonable price for it you'll likely do pretty well. Sure, other people will become immensely successful by cheating or by gaming the system in some way. But what difference does that make? Just do your own thing.
I think this is the point. Because the game is actually the psychology around why people buy shit. I can't make myself do that. For example. I can't ride the wave of hype because I don't believe its value is real or sustainable. Well a lot of excuse, right, that's why it's my trap.
Then don't. Or maybe you subconsciously don't want to do the work or have no good ideas. That's OK. But I can't see how this cynical attitude serves you otherwise.
As a reminder YC's mantra is "build something people want". There are plenty of now wealthy people on this forum that did that. Some never even had to do marketing because customers were so happy.
I don't get why they were so happy with avatar generator for example. I actually understand it's sort of fun. But $30 for that is a nope to me. This is why my mindset limits my idea space. I need to get out of that.
Read Blue Ocean by Kim & Mauborgne. What is "really good" is that what shortens the distance between the problem of the consumer and the solution.
As an example, audiophiles might sneer at consumers buying "cheap" "digital" "crap" instead of having the good sense to appreciate what a high end speaker or vinyl system is capable of, but those consumers aren't upstream, they're trying to solve the issue of needing a sound system. People in the industry are for the most part audiophiles, so they're going to have an appreciation for things their consumer base is basically illiterate in. Still, the business is run on solving problems, and the problems of most are not the problems of the high-end user.
we're solving problems, or solving really hard problems, but either way can't deceive ourselves if we're just trying to solve the problem of making ourselves money, or actually of providing value to a variable population of problem-havers.
yeah we have to blur it out... your excellence can be sold but only to a few, the rest will want various degrees or simplicity, cost etc
anything in that game requires a lot of planning to make something livable that can be sold at many groups over many years so you can finally get enough leverage to move the minds along some value axis
I think most thing sold because of lust. Example: 1) avatar generator by ai 2) my local aunties bought iPhone Pro Max just to facebook and chat .. and of course taking pictures.
It doesn't matter. But as I said, I can't make myself do it. I can't made up beuatiful words to describe pseudo values that I can't provide. For example, I don't believe in productivity app like timer that records many activity in a day, but people do really buy it I couldn't make myself believe .. like really? So when something is obviously good to sell, and I think it's dumb, I guess I can't make myself sell the thing to dumb people. Ah.. it's a trap.
>>wow thats nuts. acorns, rocks and shit. touch the dirt and rub it on your face. mushrooms are literally everywhere.
If you want to really blow your mind, start looking into soil biology. Entire ecosystems of bacteria, fungi, multicellular creatures, plants, minerals, water....
And ALL OF IT IS CRITICAL TO LIFE.
People just think "oh, it's just dirt", and move on, when nothing could be further from reality
We like to believe that the problem is just "our own adult psychosis", but I spend my days around many people over quite a range of ages (including elderly) and it is very obvious to me that cognitive abilities decline in the way shown in the charts you have probably seen (or, it feels to me, even more dramatically) and that it is apparent at an earlier age that we would like to think. Old people lose the ability to infer - if they know two Johns and I say something about one of them (just referred to as "John"), they will be confused even though it should be easy to infer which I meant. That might not be a surprise, but from my experience the same effect can be traced back linearly to much earlier ages, to those mature students in your university class, who were probably only 40, and who continually caused us to roll our eyes by asking questions with obvious answers - even at that age they had lost some of the ability we had to infer the answers from what we had already been taught. Such abilities are useful when coming up with ideas for new products, so I have no doubt that the likelihood of doing it successfully declines rapidly with age.
The example from University is not a good one. There are bright and not-so-bright young people too, but older people typically have fewer misgivings about asking "dumb" questions that may have obvious answers.
There's an interesting phenomenon where middle IQ people see bright people as rather stupid. They can't see the extra layers of complexity and possibility that bright people can, so they read "Yes but...?" questions as naive and uneducated.
Obviously I'm not suggesting all mature students are hyper-intelligent, but I do think it's not as easy to tell as it might seem to be.
> Old people lose the ability to infer - if they know two Johns and I say something about one of them (just referred to as "John"), they will be confused even though it should be easy to infer which I meant.
Old people have been bitten by the “A 95% inference will be wrong 1 in 20 times” thing a lot more, and also have been exposed to a wider array of experience where they might often judge an inference as being less certain than someone with a narrower range of experience would.
You have inferred something that may well be untrue.
A 40 year old mature student was last in school more than 20 years ago. Younger students were last in school as recently as 1 year ago. Learning is a skill and it's perfectly possible to be out of practice. There may also be some knowledge gaps because recent practices and knowledge are different to common knowledge 20 years ago.
It's also possible that mature students are more assertive about asking questions.
And also - what some of the students in a class consider "obvious" is not at all obvious to all of the class.
Of course if someone is showing clear evidence of dementia, that's a different issue.
Good communication has long been known as difficult to achieve. I could easily infer answers. But by asking questions with "obvious" answers I regularly uncover miscommunication. Try it.
If you're willing to accept innovations in place of inventions and if you're willing to consider the central message of OP's post i.e. needing a childlike sense of wonder, then...
1. "Mum, when I grow up I am going to invent a telephone with no wires that's also a flashlight, a compass, a map, a jukebox, a movie theater, and even an ENTIRE computer!"
2. "Dad, when I grow up I am going to invent a table saw that will JUST KNOW when it touches your finger and it will disappear so fast that you'll only get a little cut next time!"
3. "Sis, when I grow up I am going to figure out some way to make it possible for even six year olds like you to have their own TV channel and put whatever you want on it! You can show everyone your toys and dresses. We can call it Roxanne-TV!"
well, era of antibiotics was started with the curiosity about empty space in a petri dish.
both telephone and rubber tire vulcanization process were because of an unintended oops mistake while exploring / trying make sense of world.
adults tend to have wider interaction with world than kids, but kids dont have the baggage of years of preconceived notions/rules of adult experiences.
The fact that one needs to google the list and even then may still fail to answer is kind of the point. The struggle to give an answer is the actual answer.
This myth that children's ideas are somehow more valuable that adult's is nonsense. If this was true, businesses would organize and act on it. Parents would be hired to bring children into brainstorming sessions. Filled idea diaries would be sold and bought in bulk. Sponsored kindergarten and school networks would be established to gain access. Children hackathons with prizes would be held every day of the week. Video games would be designed to siphon ideas.
None of this happens. Because the reality is: children are great, their ideas are useless, most beautiful genius things in your life weren't invented and designed by a child or a nameless enlightened being in a tropical bungalow, but by a perfectly boring adult at their desk in their work time.
somethinggreen wrote: "... This myth that children's ideas are somehow more valuable that adult's is nonsense. ..."
And who best motiates the parents to do something / spend $$ / go in a given direction? aka commercial supplying context(s) of idea(s) to child
somethinggreen wrote: "... Because the reality is: children are great, their ideas are useless, ..."
Only until that child understands what context to put that idea(s) in to make the idea(s) usable and/or what to do / to learn to be able to use the idea in context. Adults just have a larger context node pool to draw from.
Lots of truth to this. Just start building. Build for yourself and real people, not imagined profiles of people. The more you build, the more you explore, the more you'll find low-hanging useful ideas. As you encounter them it’ll surprise you that the space is mostly empty.
Patio11 tweeted about using ChatGPT to immitate social signaling by rewriting content. I thought that was fun so I started doing it. That grew into Spencer Westford: https://vc.blankenship.io
> Lots of truth to this. Just start building. Build for yourself and real people, not imagined profiles of people.
I've been doing this for years and all it has done is drill me further down into super-niche rabbit holes that the majority of people don't care about.
I mention this because I see this kind of advice all the time on here, but it so often comes bundled with the unwritten assumption that the kinds of things the reader will build won't stray too far from the mainstream.
After more than a decade of "building", I can state with confidence that profitable ideas don't spring up from countless hours invested in building e.g. a better comb filter for NTSC.
The reality of the situation is that if your interests are things like "doing cool stuff with AI" or "SaaSify yet another thing" or "shoehorn Merkle trees into yet another square hole", then this advice will likely work for you.
If you're the kind of dude (not me) to write ld-decode¹, then it will likely not.
I mean, shit, Donald Graft² has been "building" far longer than I have and yet none of his "building" has led to a business AFAICT.
Well, actually the ChatGPT essays-fed answer (https://twitter.com/stef/status/1617222536344764416/photo/1) is pretty close to what Paul Graham himself said, although he would not admit it. Real Paul talked about anomalies, and that's eerily close to "accidents and mistakes" in my opinion. :)
Short and concise, this combines two of my most favorite mindshares.
1. "Hackers and Painters" – I stumbled across in late 2004 at the Technical University of Vienna's (excellent) bookstore. Reading it changed my life.
2. "Ideas are the Enemy of Observations (2012)" – Just Three months ago yamrzou shared this gem of a text. For me, it has become the most meaningful post ever shared on HN.
"How to get new ideas" seems to fit snuggly right in the middle :)
I read the essay and found it reasonably entertaining, but not exactly in a good way. What exactly is the writer trying to accomplish? I didn't get it, but maybe it's about having an "unbiased" look at the world, probably/maybe in the context of inferring causality, not just for lolz.
The essay writer's main problem is that they are attempting to infer causality from observation or to find an explanation of the data that justifies or supports their worldview. After all, "why?", "how curious", "I wonder how that happens" are powerful questions and statement that have led to countless inventions, investigations, novels, and pleasurable, although at times frustrating, time.
I would recommend something more scientific: observe the world (data/observations), develop a hypothesis that is hard in Platt's sense (meaning it should be discriminating) and test the hypothesis on new data and observations, where new means not used to develop the hypothesis to be tested. And they will see that they should not only collect observations that describe the world, but develop and test hypotheses that can test the mechanisms that lead to those observed facts.
Sometimes this is possible, sometimes it is not.
But it is always entertaining.
I feel I can speak to this pretty well, over the past year I've launched 130+ projects as part of https://ae.studio/same-day-skunkworks and I've come up with ~250+ ideas.
Similar to "look for anomalies" my general rule of thumb is to solve some sort of problem, even if it's rather small.
So for example:
1. I pick a topic/product. Let's say: Gmail
2. I look for things that I'd like to be able to do, wish I could do, or consider frustrations I've experienced in the past.
- I want to review how many emails I get a day on average, where are my "inbox metrics"?
- I want to see threaded messages more clearly, why is the design so opposite a normal chat message?
- I want to create emails faster, why do I have such an elaborate editor and format for something that I typically want to use informally?
3. I consider ways to solve said problems
- Where are my "inbox metrics"? - Chrome extension to show average emails per day, and graph on a timeline which types of emails I get on certain days. Huge personal data opportunity here.
- I want to see threaded messages more clearly - An app that connects to my gmail that intelligently restructures my emails to make them appear like chat messages. It handles all the noise, I just message as I would in any instant messenger.
- I want to create emails faster - predictive text templates, quick starters, GPT powered replies, change the "bulky editor" out for a simple text box + clever replies.
4. I then research if there are existing solutions to said problems, and if there are, I either ignore and move on, or consider ways that I could quickly one up their product to make mine the better product in the marketplace.
For my specific use-case, I'm trying to build a product per day, so I'll look for speedy ways to solve the problem that I've selected in a novel and agency increasing way. But I think this is a great structure for anyone looking to build a startup or a small side hustle.
250 ideas is incredible. if only 1% of them grow to have ~$1MM in value (or anything close), you should have two or three real winners by now. If only by pure chance.
I'm really curious about how that's going. I can't quite suss out what your biggest successes are on your website.
That's definitely one of the goals that I'm tracking for, though I want to stress that profitability per project isn't the main goal, so some of these projects have no immediate monetization strategy. Additionally, the projects are of a pretty strict scope (6-24 hours of work) for MVP, so we're mostly trying to improve the overall structure.
In a year I hope to have 2-6 that are decently profitable, and ideally more that are valued in the open source community.
I think a few of the most successful ones so far are:
combined web traffic is around 5-10k views per month (across the pages you see on the main site)
Side note: I'm improving that page to make finding popular ones easier. Some aren't worth more than a quick laugh/glance, while others are really worth using.
HN should have a [chat GPT] tag for stuff generated by chat GPT. So that I can ignore it. I do not mind reading what paul graham said about how to get new ideas. I do not care what chat GPT said. What chat GPT said may or may not be a good summary of paul grahams thoughts, it may be wrong in an important aspect, etc.
If Paul Graham wants to say something, he should say it directly and not hide behind a chat gpt layer.
The thing is, I have to trust that PG wrote the words saying he wrote his own response. What stops ChatGPT from saying those same words?
I'm wondering what ways/technologies we'll create to tell if someone actually generates text themselves. Right now, I can feel a rising desire to avoid text-based communication with people, almost a fear of getting catfished by text.
Yes, but is the idea sound? ChatGPT can produce fine-sounding nonsense at least as well as a human can. The difference is that I have a lot more experience calibrating my BS meter with humans than I do with ChatGPT.
Right! I think we might be the only two on this site who have more ideas than money haha.
How to get new ideas? How do I stop them coming?! My method is to scribble them on a piece of card and chuck that into a tub to get them out the way for now.
I know that tub is being emptied with nothing in there seeing the light of day. Simple fact is if I work on my own stuff I don't get to pay rent or eat this month.
No, my ideas are not compatible with VC funding, unfortunately, because my ideas only help humanity overcome our energy crisis rather than enrich some select few.
It might not be good at generating original posts, but it’s really good at generating HN comments.
Try it!
Generate typical HN comments to this PG post [ … ].
- Include both positive and negative comments.
- add replies that focus on annoying details that draw attention to the commenter
- include fan boy defence of original post
- point out obvious exceptions
When I was in my 20s, I could code for days on straight. I had so much energy and the perception of the bottomless capital of time. Now that I am forty, the cost-benefit analysis is getting tougher.
Can I have my big break and retire early? Sure, but the chances of that happening are not that great. Many serial entrepreneurs fail multiple times with nothing to show for it.
All I think about now is - when I am on my deathbed, and I look back at my life, will I regret not living it, or that I didn't spend enough time writing code in front of my computer on an off-chance that my idea works out?
Sounds like burnout. The reason you can spend days on end coding is because you're excited about the thing you're working on. If you're coding something that you don't find exciting or that is just a pipe dream, you'll have thoughts about your deathbed. That's pretty natural.
Pretty much. Every project I start, it's the same kind of yak-shaving. Database schema, logging, authentication, display of errors, build tools, CI/CD, backend, front-end. Full-time is not even enough for this for one person.
There is a ton of CRUD and mind-numbing tasks you have to churn through to get to the interesting parts, and I have been doing the boring parts for 20 years. The goals barely changed, but how we implement things has. New ways to build, new way to deploy, new languages, new frameworks. Just keeping up with it is a struggle.
I think my next project, I am just going to go Rails. It's my way of giving up :)
The "gaps that seem obvious"-notion describes exactly how I feel about pull request tooling these days. The status quo for PR review has very obvious-seeming improvements that have not been pursued (ex: de-emphasizing moved code vs. deleted code, making AI predict what comment will be left before dev starts typing, auto-reviewing trivial changes).
If I'm correct that PR review is currently much less efficient than it will be soon, it won't be because I'm smart or this is a new idea. It would just be because our company has spent last five years building code review tools, right place right time. Eventually there was enough infrastructure accumulated (and ambient events unfolding, ie OpenAI) that it became a small step to pass the edge of what PR review meant circa 2022.
I think that finding gaps as Paul Graham says is a good approach to finding ideas.
My approach to finding ideas is to think, write and use words I like to explore creatively and join arbitrary ideas together. Writing and talking IS thinking itself!
I even journal all my ideas in the open on GitHub! I share all my ideas with everyone, including startup ideas. Links are on my profile. I LOVE SHARING IDEAS.
I enjoy reading people's problems to encounter inspiration for problem solving. I would read Slashdot when growing up to learn about computer issues people were solving.
Some words I enjoy are "mesh", "parallel", "multiplex", "multi", "schedule", "marketplace", "tree", "additive", "auto", "query", "traversal", "graph".
Neuroscientist here. Closer I got to the physiology of mental health - sleep, stress, exercise, and nutrition - realized it was another fractal for personal and population health. Problem remains navigating regulatory or reimbursement hurdles from wellness into professional care.
Would you mind giving a high-level summary of the physiology of mental health (like one paragraph or two, however much you like)? Just the main things that are currently known/believed.
The autonomic nervous system and the glymphatic nervous system are both things I never really learned about in schools. ANS is taught as a textbook fact of homeostasis, not the constant fight-or-flight reflex that’s on every second of every day and the impacts that causes on health. The glymphatic system has only been discovered in the last 10 years but helps explain why sleep deficits explain cognitive performance decrements, mental health concerns, and age-related declines. And still so much we don’t know like gut-brain health and neuroelectronics as for implants!
Have a anomoulus brain, that recombines things that do not belong together (conspiracy theories) but have enough filter to recognize the the bad recombinations. Train said brain with rewards for recombinations, even ridculous ones.
Also recongize that inventors would be excellent in story writting when asked to produce a recipe for creativity. So the step by step instructions may be nonsense.
Creativity, the far reaching one, not the incremental one, is topological adjunct to insanity. Which is why the radical new things, only get funded in times of desperation. If I were a inventor, i would be heading for ukraine or russia right now.
One of the oldest saying in entrepreneurship is "need is the mother of invention"
It starts with finding the right problem/need and before jumping into to solutioning, the smart person would validate and pitch the problem to as many relevant target customers as possible. Once the problem is validated and tangible root cause is flushed out then it's time to explore the solution and pitching it to the same customer research group for validation before going into building mode.
Finding ideas is not hard, but the process is gruelling, frustrating and potentially rewarding. Good luck!
I thought GPT wrote what Paul posted! I guess I cannot tell the difference anymore between who's saying what. I actually thought what Paul wrote sounded like a poor imitation of Paul Graham... spooky!
And I'm over here trying turn off my endless torrent of new ideas, to better focus on the work in front of me.
During the apocalyse, it got really bad. Probably due to isolation and so I didn't have my usual inhibitors. ("Being polite company" turn upkeep.)
--
The one strategy I have "to let it go" for my idea factory is to journal. Write however much of the notion down in notebooks that emotionally allows me to mentally stop chewing on it. Sometimes just a one liner. Sometimes long white papers.
Real progress in any field or endeavor comes down to about 5% new ideas and creativity, 95% hard work putting those new ideas into practice or building something functional out of them.
That 5% is pretty critical however, as without it you're just continuing to redo what's already been done.
The problem with using percentages in these contexts is that they refer to something seemingly intuitive, but not quantifiable, and therefore ambiguous at best and useless at worst.
What does 5%-95% refer to? Time? Effort/energy? We don't know.
It is like saying that to lose weight 90% is about nutrition and 10% is about exercise. 90-10% of what, exactly? And what would change in my approach to diet/exercise if it was 80-20% or 70-30%?
I think vague quantification does more damage along the way to achieving goals than we suspect. Maybe 10% more.
I don't think it's easy at all. Try to come up with an original plot for a hollywood big budget time travel movie. Take the plot of Edge of Tomorrow. I could never come up with that. Ideas are easy when it's just "Grand theft auto in space", but actual originality? Extremely difficult.
On the other hand I ditch a handful of film/TV ideas each week.
Many because they'd be crap of course, but the rest because there's no way in this lifetime I'll have the resources to bother thinking about them, never mind have the time to actually write them out.
Gotta do regular people work till I die. So glad I was dragged out of the universal energy just to pay rent for 60 years haha. Fuck I wish I could afford to do something arty farty instead. Such is life.
I passed my GCSE English literature (end of high school exam in UK) writing a time travel story, haha.
That expression, "ideas are easy, implementations are hard", is lacking nuance. No, new ideas are ridiculously hard: can you have the idea for an alphabet if you have never seen one before?, even the banal equals sign is a ridiculously hard idea, Euclid and al-Khwarizmi never had such an idea [1]. Ideas are hard, implementations are sometimes hard, idea collages or idea twistings are almost always easy, given a ripening context: sliced bread; sell things, but on the internet.
I think the key is noticing, which implies being a more passive observer. Most people want to find new ideas, which is being active. I've never had success at actively finding ideas but much more successful just living life, listening and the ideas just come.
It’s not about whose terms you’re living on, it’s about listening. Most people around us are dealing with mundane things that need fixing, just listening to people’s grievances is a great way to ideate. There are individuals who you just end up meeting through life that have great niche problems that require nothing else other than living and listening to attain them.
Going out and “pursuing” ideas often leads to some confirmation bias as we’re determined to “find” an idea, rather than sitting back and letting the lightbulb go off when you’re not actively trying.
but the invariant of checking out/exploring just for the sake of being somewhere but 'here' does present one with way more opportunities to listen/hear.
i am genuinely surprised that pg is writing such a short response to a question that seems right down his alley, seemingly very core to his subject matter. how has he not written an authoritative “how to get ideas” post until now?
I think that's why a lot of innovation comes from "outsiders" rather than industry veterans. You are more likely to notice the anomalies with fresh eyes and stop noticing them through time.
Conversely, I have seen that some of the most boring startups that you never hear of, but are killing it, are built by domain experts who know the task they're trying to solve inside and out. Two examples from my own country, Spain:
- A man who built a SaaS for local administrations and has about a ~50% share of the market. He spent 30 years working for the Spanish IRS in different positions. I don't know the insider story, but looking at the website he just made a SaaS solution with good UX in a space where UX tends to be atrocious.
- RatedPower's CTO. RatedPower makes a software that automates the design of renewable power plants. This guy had been drawing the plans by hand as a renewables engineer, realized a computer would be perfectly equipped to do it, and the competition in the space had grown complacent and barely innovated in years, so he did it.
The older I grow the more I get excited by boring ideas. People pitch ideas to me frequently and most of the time I literally end up thinking "not boring enough"!
I agree. Things fall into existing grooves. Your band comes up with a new sound and people say, "That's kind of like Mazzy Star," and now you point it more that way. It falls into that nearby groove instead of continuing on its own. A lot of great sounds came out of the 1960s because there were so few grooves. No one knew what the rules were.
Plus, a lot of low hanging fruit stays there because no one looks at it. Everyone is looking at x, y, and z, and there are a thousand variations of those, and they ignore a, b, and c because no one brought it up.
Yep. Or somehow focus more, that would help knock a few more ideas out as well.
My backlog of games/puzzles/stories I'd eventually like to get out there might be >100 at this point.
Focusing mostly on a bit more of an ambitious game right now, and it's probably still 6 months before I get it into Steam early access (coding in my spare time and energy, which is sparse lately), so everything else is just waiting at this point.
But sometimes I get an idea that I need to at least get some playable prototype with index cards, pen, paper, and tokens, at least. Just had another concept pop into my head a couple days ago for a roll-and-write I should at least make a board and write up some rules for before I forget.
Waiting for the day when something like ChatGPT can code up a basic but working game based on my description of its mechanisms. Even if it can only handle abstract games on a sqauare or hex grid, that'd still be useful for me.
I am interested in that? Maybe we can use ChatGPT or related tech to change Dungeon Crawler games or related procedural generated games? I think DCSS would benefit from this, imagine designing entire levels so that they fit your character well. A common frustration between people that play Felid, draconians or Octopodes according to the dungeon crawl subreddit is that, a lot of armor that generate that they can't use. I think if we had better code or algorithms that targetted even something as basic as item generation based on class or race you play will make games more immersive for people that pick them. It's always weird when I play DCSS, and I am playing a human brigand or something and I see Centaur warding or armor that I can't wear. Some of the best games I have had are where the character that I play, gets items or environments or vaults that help my character progress, this means the correct spells, items, stat increases, enemies and so on.
After I hit midlife I decided that I just didn't have enough time to do them all, and started pruning aggressively. I now keep three projects alive and bounce between them as interest directs. It takes discipline. But it also exercises my focus which was never a strength.
I'm mainly working on one video game and one board game design at the moment, mostly the video game right now. Sometimes if I want a break from the game I'll do a tiny bit of work on a library that should facilitate future web game development, or trying to get a personal site up again to showcase all my games (including past ones that have been released and videos for board game prototypes I've pitched to publishers before).
I do have a list of what to work on next after that, though. I did used to try to juggle a lot more and bounce amongst several projects at once, but it's been too long since I've released something, so I've got to knock it off for a little while. But if I don't scratch the creative itch when it strikes at least as far as some notes and some card ideas or something, then I start feeling pretty stuck, since a lot of my current game is just getting all the UI running and polished right now, not exactly super creative stuff.
While I do have a VR device I've done the whole digital thing before (I have versions of some of my games on Tabletop Simulator, in fact one of them is downloadable on Steam Workshop right now) and it's just not the same as being able to handle things physically.
My brain doesn't catch things as easily without being able to physically manipulate objects. Like one example is I didn't notice how fiddly resource manipulation was when I only had a digital version of one of my games. It was only when I was manipulating it physically that I went "yeah that's not going to work".
Without playing with the physical bits, it's harder (for me) to come up with special actions or different ideas for mechanics. Not impossible, just harder.
Also my local playtest group and game designer conventions meet in person, so I better have something physical for those.
I do write simulation programs for some things though, like to help balance scores, determine if there's a first person advantage, how often do draws happen, etc.
However, I am interested in the recent push for sustainable game production, with people making roll and write games where the sheets you play on can be done on an iPad without printing anything out. I may have a design that's compatible with that, a concept that popped in my head recently that I'll start designing soon. One example[1] and another [2] and another [3] of these games on Kickstarter.
I reuse a lot of components anyway. I have a tub for wooden cubes, another for wooden meeples, a third for dice, a fourth for money chips, etc. and I have a few decks of cards that have six suits and numbers from 0 to 15 on them that covers a surprisingly wide range of games, that I can just grab and start using. But if I want to have special cards, then I use index cards (and eventually do graphic design and print card art for).
Also not sure why you think VR and E-ink is somehow not a consumable resource. They take an incredible amount of resources to make in the first place, probably far more than some sheets of paper, and eventually they'll need to be upgraded or replaced, which uses much more resources as well.
The issue is having a full time job, leaving little time where I have creative energy to work on side projects. Strongly considering trying to find a part time consulting gig once my current is played out.
> Someone fed my essays into GPT to make something that could answer questions based on them, then asked it where good ideas come from. The answer was ok, but not what I would have said. So I asked: What would I have said?
It's not entirely clear whether or not the essay was written by GPT because it's not clear who PG asked. Himself or GPT?
I'm leaning toward the former, but can't rule out the latter. Maybe that's the point.
not an expert, but i think usually one of two ways - either use openai’s finetune api (the right way) or embed all the docs, retrieve them using embedding of the questions, and add just enough of them to the prompt to generate a response (gpt-index or langchain help automate some of this).
downside of former is lack of sources, downside of latter is context constraint
currently, search engines/chatgpt seem to fill that roll vs more historical library/university.
aka search term "frontiers of knowledge" <replace with subject area field(s) unfamiliar with>
compared with traditional library/university book writer/teaching assistant
note: sed & awk not required to do text replacement, but good editor certainly helps
But if patient enough to wait, will eventually show up on internet as a video(s) way before the random monkeys typing away at a type writer get to it [1]
TBH, at this point, if there is a problem that hasn't been solved, its because some people tried and failed and found that the problem is not solvable currently. It may be more fruitful to start with something promisingly popular, and improve upon it (e.g. tiktok)
This can work. I've done this for some of my game designs. "ChatGPT, I have a game design with this theme and these mechanics. What are some ideas for some new special tiles / cards I can add to the game?" And it'll give me 5-10 ideas, some junk, but a few that are worth exploring and I might not have thought of on my own.
I'm finding the whole AI thing discouraging. Nothing like earlier tech gold-rushes. Every single idea I have starts with "have, or acquire, an absolute shitload of capital"—because you need huge, tagged datasets to move forward with any of them. That means capital, and lots of it. And if you find a way around that you're still burning non-trivial amounts of money on every little experiment (for compute time) all in search of an outcome that may not materialize, i.e. you may never even have anything that you could attempt to sell or raise money with. Seems impossibly expensive to engage in just to fail several times, unless you've got a large pile of capital.
It’s actually so short, you can read it here. (If you’re like me you head to the the comments section before going to a page…).
Here it is:
How to Get New Ideas
January 2023
(Someone fed my essays into GPT to make something that could answer questions based on them, then asked it where good ideas come from. The answer was ok, but not what I would have said. So I asked: What would I have said?)
The way to get new ideas is to notice anomalies: what seems strange, or missing, or broken? You can see anomalies in everyday life (much of standup comedy is based on this), but the best place to look for them is at the frontiers of knowledge. Knowledge grows fractally. From a distance its edges look smooth, but when you learn enough to get close to one, you'll notice it's full of gaps. These gaps will seem obvious; it will seem inexplicable that no one has tried x or wondered about y. In the best case, exploring such gaps yields whole new fractal buds.
though moving away from pg rating to statistical r, might result in some twisted python stuff. especially if ast taken to the limitless turing resources extreme.
One of the surest ways to come up with new ideas is to pay attention to what doesn't fit in. We're used to seeing things in a certain way, and often times it's the things that don't quite fit into our preconceived notions that can provide the best insights. The most obvious place to look for these anomalies is at the frontiers of knowledge.
Knowledge grows like a fractal; from a distance its edges may appear smooth, but when you get closer you begin to see the jagged gaps and spaces. These gaps can be surprisingly obvious; it can seem strange that no one has thought to ask a certain question or investigate a certain problem. When you explore them, you can often discover entirely new areas of knowledge that have been previously untouched.
That's why it's useful to investigate these gaps in knowledge. It may feel uncomfortable at first; we don't always like to challenge our preconceptions and explore things beyond our comfort zone. But often times this can be the source of some of our best ideas.
You don't even have to look far for these anomalies; they can be found in everyday life. Much of stand-up comedy is based on this, by finding the oddities and quirks about our daily lives that we normally take for granted.
So don't be afraid to take a closer look at what seems strange or missing or broken. It could lead you to unexpected discovery and perhaps even uncover some of your best ideas. This may be challenging and uncomfortable, but it's often worth it.
per 'doesnr sem to fit in': is looking outside of ones field and/or historical anacronisms reframed to current day/technology cheating when comes to idea developement/gemeration?
Me too. Most blog posts have one pretty simple idea that it takes a ton of reading to figure out what it is. I appreciate the time he put into distilling it.
If he were to make it longer some examples might be interesting to read about and drive home the message, he no doubt has examples in his mind, and his readers would no doubt be curious as to what they are
For example, laptops and phones produced 20 years of incremental improvements in li-ion batteries, and suddenly electric cars are "unlocked".