How I won $2,750 using JavaScript, AI, and a can of WD-40
davekiss.com
davekiss.com
Better than the piece. Better than the essay about the piece. They want the summation of the essay of the piece. Preferably with an associated pic of a cute art goth pixie.
Ah art.
My dad is an accomplished artist but never wanted to play the art scene and mostly spent his career as a professor. He once produced an entire storage shed of metal sculpture than he later recycled for lack of anything better to do with them. I could only fit two in my house.
Also, I don't think you have to censor words here in general.
We’re an adhocracy that uses votes, and profanity doesn’t win votes…simple math…
This requires a serious mastery of the art that allows to obtain exactly what's envisioned.
It also rubs many artists the wrong way, because they want to be led by the art, ideas, concepts, etc, and not by external riders. OTOH most great masters of the past worked under such conditions, with notable success. Michelangelo certainly had a rather detailed list of requirements for the Sistine Chapel paint job.
From what you do at work or say to your friends to whether you break a writing rule or whether you jump a fence.
When you observe people in their element, you can tell apart the people who know know when to break rules and who haven’t figured it out.
The rubric also put a ton of points on the technical report which seemed like an after thought for most people but at the end of the day it was an engineering competition so the whole point was the technical portion and the dog house display itself was kind of just the fun bit. We did really well just because we bothered to do all the requested calculations on life cycle cost analysis, heat losses, carbon balance, etc.
Our final product looked like a tank, but it won by a long shot.
There's a solid educational consensus on how to become the best at something: do it a lot.
If your goal is to winning contests, you should really prioritize cranking through them: getting practice at getting better. Don't try to make a masterpiece. Instead focus on becoming fast at creating entries, doing them according to the rules, and sending them off. Like those studies where people who made 1,000 good-enough vases were better at the average one than people who tried to make 100 beautiful ones - even when the goal was 'create a beautiful vase'.
But then, if you're a professional contest-enterer, that's got do a lot to boost your chances - especially over time, right? At the 1 year mark, I would expect you could raise your chances to 10% of a win of some type per contest (being conservative here). You've specialized in this. You'll always be a contender.
And that number's high enough that, if you enter 10 a day, that seems closer to the optimum. You are playing a numbers game on the strength of your numbers.
By analogy, this is like picking an index fund and continuously investing, rather than trying to pick that one stock that's going to the moon.
What I took from this was:
1) Choose contests with well defined judging criteria, not subjective or popularity contests. Ideally ones where there are specific weightings for requirements and where some of the heaviest weightings are both a) things that'll be overlooked by a lot of entrants, and b) things you're personally very good at. (in this case video content and production)
2) Choose contests where there are sufficient prizes that it's worthwhile entering even if you don't win first place.
3) Choose contests where you are allowed multiple entries, and allowed to win multiple prizes.
This is all backed up with some javascript, webscraping, AI, and math - presumably to meet all the judging criteria and improve their chances of winning the HN front page contest. My suspicious is that this is what most of that validation you're skeptical about was for - not a beautiful vase, but a checkbox ticking exercise in ranking on HN.
Then don't bust a gut trying to produce the best possible entry, just use your subject matter expertise (in this case video content production) to produce "good enough" entries that are above the level that most of the public can create but without striving to surpass Ridley Scott or Quentin Tarantino levels of production.
Don't "make 1000 vases" to practice contest entry and enter 10 contests a day. Instead carefully choose only the contests who's requirements include something important that you've already "made 1000 vases" for (like, whatever you do as a day job) and only bother entering those.
Particularly as there's no reputational downside to entering and failing at such a contest.
The odds were calculated up front.
It's better to enter a winnable completion rather than one that cannot be found without disproportionate effort.
Hackathons are a good example. Global, online ones tend to be long, time consuming and near impossible to win any prize. Compare that with local only, now there's a chance as the number of competitors is limited to the geographical region.
I remember wrapping up the competition, winning all the prices we could win except one (1st place, and "best of" for 3 out of 4 categories), and the feeling that we had demolished the competition all around and that felt somehow a bit unfair for the rest. That was the last time I participated in a competition of this kind.
How did we do it? I had the network of contacts already (co-created a Maker group at my University) so I could literally hand-pick a team of 5 with the best maker in each sector; I called my previous co-winner who is great at tech presentations, and called the best designer+3D artist, the best App programmer, the best electronics person I knew, (and me). I got a stomach infection and spent a third of the competition in the toilet, but I ended up doing mainly programming and organizing: 1-2 hours to decide on the theme and solution, then split the work efficiently, and then let each person work on their own thing while continuously checking for sync between them. We ended up with a product and UI that seemed like a commercial product.
If you're a professional contest-enterer then it seems like a pretty bad idea to produce a detailed write-up which prospective rival contest-enterers may use as training material. If 1k people from HN visit your blog then even 0.5% of readers getting in on the game can cause your expected value to plummet.
It should be noted that the article addressed the question of how to discover contests with a refusal to divulge. This is security by obscurity. If a bunch of people get excited enough about entering contests using these deliberate winning methods then it wouldn't take much for them to set up a Discord channel to coordinate the discovery process. If even 10 people get together to "entry bomb" these contests then the original author's advantage will evaporate.
We're talking about a national newspaper with a circulation of ~3 million at the time. I still keep the copy of the issue with my name. Sometimes, just showing up can take you to the top 0.0002%. :)
A British computer TV show ran a content on their web site, but it was a fast-paced multiple-choice Flash game. I just opened three accounts. Ran through it with the first two to figure out all the right answers, then got perfect scores on the third.
The prize was to co-present the show one time, but apparently they got so much positive feedback they made me permanent, until I found a new job. I never let them know I cheated :/
This is the show. I only did it because I had a huge crush on Kate Russell:
> I've got a Voodoo Banshee graphics card and when I installed Win98 with the latest drivers meant I could play the new titles like Unreal Tournament and Quake III but couldn't play old titles like Rollcage Turok 2 and Quake II. I then decided to install Win98 Second Edition with my old drivers which worked so I can now play all my old games, however, I've got no Open GL so I still can't play Quake III. Please give me an answer!
> Paul
The nostalgia is hitting hard here.
This appears to be true of writing anthologies; it's pretty common that Amazon is running a sale on some anthology or other that advertises an incredibly high-powered list of contributing authors. But the quality on these is lower than you might expect.
I understood why when one of Brandon Sanderson's blog posts mentioned that he was solicited to contribute to one of these that was raising money for some charitable purpose, and he sent them a piece of writing he'd done for a different book that hadn't made the cut to be included in that book.
You attack the problem like blackjack card-counter would. You assess the rules, make mathematical odds projections when possible and logical ones when not, and keep a keen eye on what you are up against as to judge how to best attack the money.
Thanks for the smart write-up...its been a big inspiration for me.
Lubricating a sliding door
Quieting a squeaky door hinge
Lubricating an HVAC register lever
> While the “W-D” in WD-40 stands for Water Displacement, WD-40 Multi-Use Product is a unique, special blend of lubricants. The product’s formulation also contains anti-corrosion agents and ingredients for penetration, water displacement and soil removal.
Regardless of the facts, when all was said and done, they were the ones judging the entries, sooo… yep, it's a lubricant.
I once replaced a series of mortice lever locks in a first house that were very worn after a decade of use. You had to shake the key to get them to open. I was amazed on opening them to find them completely dry, when they come greased. I suggested they lubricate the replacements occasionally. They replied that they sprayed them regularly with wd40. This has washed the grease out and left... virtually nothing.
I've heard a lot of anecdotes in every direction: only use graphite, only use a cleaning solvent. I've never read anything authoritative and I suspect the ideal approach is to regularly (e.g. every decade) disassemble the lock, clean it thoroughly, and inject grease, but nobody is going to do that.
Still unclear on why this particular mix is optimal for unsticking things but Google suggests a lot of people use it.
You can see a similar effect with standard skate bearings used in fidget spinners - most bearings come pre-greased with a fairly high viscosity lubricant. If you spin the spinner, it will slow down quite fast, but if you take the bearing apart, clean out all the grease, and reassemble it, it will run wicked fast and long. Until some dirt gets in, at which point, grease would have worked better.
And yet, it lubricates! I've heard your claim a lot, but the fact remains, it makes machinery work more smoothly, by reducing friction, and it is effective over significant time periods. By any definition, it is a lubricant.
https://www.wd40.com/myths-legends-fun-facts/
> Myth: WD-40 Multi-Use Product is not really a lubricant.
> Fact: While the “W-D” in WD-40 stands for Water Displacement, WD-40 Multi-Use Product is a unique, special blend of lubricants. The product’s formulation also contains anti-corrosion agents and ingredients for penetration, water displacement and soil removal.
The control, ie uncoated piece of steel fared better results that the regular WD40. I remember the best being Clenzoil (I had to grab my can to remember this). However, among the best was.... WD-40 specialist, specifically the "corrosion inhibitor" version. I think it was either the second or third best and I consulted my other can to remember this.
Sadly, my favorite lube ranked very poorly, which was Balistol; however, I'll never give up my Balistol.
I have put both the Clenzoil and Specialist to various 'tests' over the years and can vouch for their quality. But I'd use snot before regular WD-40 unless I was making a stink bomb.
Edit: While not the test mentioned, ProjectFarm (youtoob), who does myriad high quality evaluations, did test various lubes, but I think mostly for lubricity. It was also revealing and I highly recommend it and the channel in general.
It's just marketing.
https://www.wd40.com/products/silicone-lubricant/
You could probably burn WD-40 in a propane stove, but that doesn't mean it's a fuel.
seems like a bit of a straw man, no?
> a silicone version that will make a sticky old lock work like new.
In my apartment, my deadbolt was so sticky that it was hard to turn even with the door half open. Two spritzes (keyhole and the bolt) with WD-40 classic, and it moved easily, and continued to turn easily for the remaining 6 months in my time there. Could lubricant-X have done a "better" job? Maybe, not that it mattered in practice.
If NASA designed an even better lubricant, would lubricant-X no longer be a lubricant at all by your standards? Because now there's something that can also unstick your lock, but with an even lower coefficient of friction?
WD-40 contains lubricants, and it can be used as a general-purpose around-the-house lubricant. Proven by my years of personal experience. Trying to claim that it isn't a lubricant is like trying to convince me that my lightbulbs are dim. I just laugh and move on.
It was likely the solvent properties of WD-40 that helped. It dissolved the old, tacky lubricant or rust and left you with a cleaner lock. For a six month fix on a rental, that is probably fine. For a long term fix on a house you own, it may be worth using one of a plethora of lubricants designed for that specific use case (including a more specific variant of WD-40 that is designed for locks and leave behind a dry lubricant when it evaporates.)
> Because now there's something that can also unstick your lock, but with an even lower coefficient of friction?
WD-40 is decent at removing rust, but not great at repelling water and thus preventing rust.
There are a lot of things that go into picking the right lubricant for a specific application. It isn't just "which lubricant is slipperiest?".
WD-40 is indeed a lubricant, but much of the benefits of using it are due to its solvent properties and people often don't understand that applying it in the wrong circumstances can lead to removing the correct lubricant and result in less lubrication.
This is either useful and appropriate for a situation, or it's not. Also, the oil might not be a specialized oil for any particular use, but a lot of applications don't need a high performance oil.
In my view, the drawbacks are:
1. Controlling overspray. This is why, even if I liked the stuff, I'd prefer to apply it with an eyedropper in many if not most cases.
2. General ignorance about lubrication needs, where something else is preferable, such as grease, a suspension of wax, penetrating oil, etc.
On the other hand, keeping a supply of every possible lubricant can be a storage problem, and I've gradually come to prefer using the "wrong" stuff than buying yet another oversized container of something that I have to keep forever or dispose of.
The ElevenLabs narrator did say the faeries applied “Polish”. From Gdańsk or Warsaw? :)
I enjoyed the post, I appreciate a good methodical process.
So the judges may have seen something different rather than protecting, preserving and prioritising your family and the joyful and creative structures for children's play. At the very least it would give ambiguity in today's more ecological minded world.
not really the same thing as spraying a bit of WD40 on a dead stump.
I did a contest once and won $2500. In the end the part I spent 2 weeks on made like $50, and the part I spent 30 minutes on right before the deadline made $2000. Not sure what lesson to take from that...
It was way more work than expected, especially because they had us do a peer review that took like 80% of the actual time. So in the end I probably made close to minimum wage and most of it was not fun. But the novelty at least made it sorta-worth it.
> here are the rules for a video contest. are there any gotchas or clauses that could give me the slight edge in winning if i were to submit one or more videos? what do i need to know? [entire ruleset]
The response was accurate but generic. Maybe with some stronger prompting (ie. "what's the number one thing I should prioritize?") it'd be more helpful, but for now, I'd still rather review manually.
See the response here: https://gist.github.com/davekiss/10209469246dde21b2550747add...
For lubrication, use a more appropriate lubricant like lithium or graphite grease, motor oil, cutting oil, or dry lubricant in dusty environments.
For a penetrant oil, 50-50 (by volume) acetone & ATF is one of the best for rusty ferric materials. (And, in general, you should be using an anti-seize compound orthreadlocker, and/or surface coatings to prevent rust.)
And apparently a shingle sealer too.
Is WD-40 a lubricant? Apparently, again, it depends!
The funny thing is where I live you can buy “WD40 Lubricant” (silicone).
I think many people buy WD40 because, unconsciously or not, the smell brings back memories.
Let me explain. I ride motorbikes. You don't want to be using WD-40 on the chain as it will strip it of any existing lubricant and dry it out, which as you can imagine, is bad. However, it's actually really useful for cleaning parts of the bike that have been covered in gunk and grease.
I have to admire their marketing, as like you mentioned, I grew up in a house where WD40 was used for absolutely anything sticking or squeeking, and so for the longest time also used it to fix those things, and beyond.
If you asked me to name the part of a bike most likely to be covered in gunk and grease, I'd say the chain. ;p
Yeah...
I love WD-40 and use it all the time but... For many use cases there are better solutions out there. For example the picture showing it used in a window (or door?) seal: you want to use a spray of silicone for that instead. For locks (inside the lock I mean) you want to use graphite powder (or teflon). Not WD-40.
WD-40 is amazing because one can can save your day. But the examples where something better exist are numerous.
[1] https://variety.com/vip/content-owner-lawsuits-against-ai-co...
Did you spend more time working on this blog post than on the contest?
Yes.
This story wouldn't have happened without AI. I genuinely used the tools to tell the story I wanted to tell.
There was something about the way friction was removed with the typical narration round-trips that actually made the storytelling better, easier, less painful… and I truly don't believe anything authentic was lost in that process.
My hope is tools like these can make storytelling easier for everyone.
I'm particularly intrigued by the creative things that can be accomplished with mutual diligence between user and AI. As a bit of a moron, I can now easily accomplish what otherwise would be too daunting. It's great..., when it works. What a smart person can accomplish under such conditions is frightening, in bith good and ungood ways. I'm hoping to go the API route soon and move toward a more free-thinking version geared as a research assistant.
Anyway, I'd say that's some pretty clever teamwork. Well done and thanks for sharing. Just get some better slippery sauce and if you're gonna squirt it on all beneath the heavens, go with Ballistol (I misspelled it in my other comment) - the fairies will thank ye.
I don't think that's the optimistic take you present it to be. To me the post reads as the story of someone for whom art is the least important part in an art contest and treats the artistic part as an afterthought ("I just wanted to adhere to recording content that prioritized the weighted breakdown").
I'm not going to pretend that the contest attracted the top artistic minds of this generation, but anyone who actually put care into the art part has lost to someone who made an art pipeline. Extrapolate to the art world and suddenly there's less real art because the money out there to fund it is going to AI programmers checking a spreadsheet.
Heavily artistic, thoughtful, and cinematic, leaning into my passion for the art of telling stories through video that I've kept close for almost 20 years now.
It lost. The judges selected the entries they felt were the best fit. The argument for this contest is that AI enabled _more_ thoughtful art to exist, not less.
I didn't particularly like any of the videos, even the fairy one, no offence. This probably speaks more to the nature of the competition than the videos themselves (I could care less about WD-40, though I'm sure a really good video could sweep me away, similarly to how I'm often genuinely impressed by the creativity of superbowl commercials, even for products I don't care about).
But I can definitely understand the thought process and also liked your write-up. My compliments to your art, which is your writing and your description of your thought process, and your creative solution to a challenge within an economic framework where you need money to survive and can justify outsourcing other typically human endeavors to the robots in pursuit of financial compensation.
But that economic framework, the incentives, and the replacement of that human touch with AI output is depressing, and the thought that most things we encounter online and in media will be AI-generated in the future is terrifying. That's not a commentary on what you did, and again, you also wrote an excellent article about it which I highly doubt was mostly written by AI.
As you said (perhaps in another comment) you spent more time on the article than the actual competition. That effort shows. And it stands out among a sea of low-effort, AI-generated blog content even now on the Internet. But when you talk about "AI making storytelling easier" there's also an admission that we're moving to a world in which, in many situations, lower-effort content and AI-generated content are becoming more worthwhile under the reality where maximizing the ratio of monetary reward to time investment is necessary and expected.
Basically, there's a lot of AI enshittification going on, but this isn't one of them. Just seems like a bit of fun.
But the author of this post saved a lot of time in the interest of dominating the competition by using AI instead.
While I can appreciate that the process of assembly and curation of AI-produced components is an art (much as collage is an art form), I'm still dismayed by how AI 'art' is replacing real art (as the components of this collage, if you will).
To me, the difference between real art and AI output is human intent. There's a context, a though, a mood. What AI produces can't be art in the same sense, because there's no human behind it, who can explain their reasoning for why they used a specific stroke in one place. And real-world incentives diminish the value of the human element, which is why we're heading to a "dead internet" fast.
But for every contest he enters and puts effort in what is the overall payoff, seems unlikely to be this high, and with that I would think it seems like a bunch of stuff to do for little money.
I have considered doing this before, contests etc. but I'm not sure I can breakdown the actual cost of doing it, the benefit, and risk to be able to see what one actually gets from the time devoted.
Not to say that I might not do other things that cannot be quantified easily either, but in the contests case I want more quantification, I guess it just seems more likely a waste of time to me.
"There's something almost poetic about working with your hands. In a world where everything has gone digital, where things get solved with a click and a swipe, there's a unique satisfaction about tackling something tangible, something real."
Then there's this guy who algorithmically floods the contest with AI slop. Click and a swipe indeed!
I wrote the script, I shot the video, I edited everything — the only place AI was even used in the submissions was to alleviate the pain of tracking down fitting music and re-recording voice overs (both of which I've done manually in the past).
I actually thought the videos would be pretty helpful for a non-handy homeowner.
But that's not what happened. The blog is very clear about explaining where and how it is used.
I know there's AI in the title but it's pretty lazy and frank rude to assume that means "mass produced nonsense".
Is there somewhere that contests like these are tracked?
Would be awesome to browse.
Where do you find these exceedingly winnable contests?
Sorry, that’s a trade secret — but it’s not any of the standard various contest aggregators out there.I found the parts I was interested to be wonderful.
I know it feels like you're 'just' sharing an opinion (or, lord help us, a 'fact'), but if you consider the externalities—i.e. the effect on the culture when many people post this way—the expected value of posts like this is super negative for everybody—including yourself—since over time it makes the commons nasty and lame. That's why we try to avoid them here.
Oil displaces water, who knew?
> Where do you find these exceedingly winnable contests? Sorry, that’s a trade secret — but it’s not any of the standard various contest aggregators out there.
I’m going to guess that if “You cannot replicate this” were at the top of the ad fewer people would spend their time reading through it.
0 (The link is towards the bottom, right above the reveal) https://davekiss.com/consult
Not in the context of the post. Regardless of what it is, it's not required to be "useful". Just enjoying the content, for whatever reason, whatever its intent, has a "use" to some.
But generally I agree. I won an Instructables contest once, because the odds were very good: most entries were low effort and it was panel judged. Nowadays I dunno, there are so many entries and a lot of them are fluff from companies writing instructions on how to build their kits. But I do think if you had nothing to do, and a good workshop, you could probably farm contests there.
"there is no limit on the number of entries" "official rules" 2024
There are no excellent examples of anything. I feel cheated.