This package has saved me so many hours of tedious gruntwork. It's like a junior developer - you still have to manually check their work, but when it's correct, it's a great productivity improvement.
And don't forget where this will go in a couple years with improved models and more computing power, it's gonna be awesome!
[/i]
1k · 2h ago
This exactly. It is more important to move fast. Screw the edge cases. As long as it’s correct _most_ of the time, you can always fix anything that’s broken tomorrow.
whatever1 · 45m ago
Why on earth did the software engineering interviews were checking candidates’ ability to think about edge cases? Clearly management does not care.
pavel_lishin · 1h ago
It's called "eventual consistency".
sho_hn · 2h ago
Developers who aren't using it are already falling behind.
SamBam · 2h ago
Remember, it's not that AI that will take your job, it's the developers who need an AI to tell them if a number is even that will take your job.
darepublic · 2h ago
pro tip: play around with the temperature especially when using big numbers as input
hzambo · 2h ago
Wow, amazing tip. This hack improved my workflow by 10x.
Waterluvian · 1h ago
Only 9x for me. What am I doing wrong? Can you share your vscode colour scheme file?
floren · 12m ago
What model were you using? You need to use gpt-3.14-tastesgreat-lessfilling, I've used it to write 130 side hustle projects this month with only prompting.
koakuma-chan · 2h ago
Are there actually still Junior Developers out there? I thought no one is hiring Junior Developers.
ArthurStacks · 34m ago
Its actually the opposite of how people think.
We hire junior devs, but not senior, and dont replace our senior devs. So our developer base is moving towards being junior weighted with less senior. The reason being a junior dev is cheaper, complains less, is more capable now through utilising generative AI, works harder to impress knowing they arent in a safe position, and we can let them go more easily with less process and less reasons needed to be given.
echelon · 2h ago
You laugh now, but our jobs are going to be toast in 10 years.
I thought self-driving would never happen, and now it's here.
iamthemonster · 2h ago
This is pretty useless to be honest. It's good for telling whether a number is even, but in our industry we need more powerful functionality. We also need to know whether a number is odd.
charles_f · 2h ago
RTFM, it's not only implementing isOdd but also a large set of rarely used advance operations such as isEqual, or isGreaterThan
Trully AI is astonishing
whynotmaybe · 2h ago
Why a specific function ?
With a few lines of code, you can just create a list with all the numbers that are even and when you need to check if a number is odd, you simply have to check if it's in the list.
turnsout · 1h ago
Yes, this is what we do as a RAG workflow. We created a list of all 32bit unsigned integers and whether they were even or odd, and we pass that into the context. The future is amazing!
MikeTheGreat · 17m ago
I'm new to RAG and have a question: how do you get all the numbers into the context window?
Does the RAG part look up just the needed number?
I think that Gemini has a million token window (yes?) - do you have access to a model with a larger window?
Regardless, I find your ideas intriguing and wish to subscribe to your Substack.
Charon77 · 36m ago
I.. can't tell if you're joking or not. Pretty sure someone out there is unironically doing something as stupid as this in production
sureIy · 2h ago
Do we have enough spaces in the ALL_NUMBERS array or do we want to group them by thousands?
ALL_NUMBERS_00001
ALL_NUMBERS_00002
iamthemonster · 1h ago
we fired all our junior devs so we can't write code any more
briansm · 30m ago
"This is pretty useless to be honest."
I remember saying that about Bitcoin 15 years ago.
crubier · 2h ago
You can achieve this super simply this by prompting the OpenAI API to call this tool and reverse the output.
SamBam · 2h ago
I tried that, but I kept getting "eurt" and "eslaf" and I'm not sure what to do with those. Do I need to send it back to the AI?
YokoZar · 2h ago
Great news! The package includes an isOdd function as well!
Cerium · 2h ago
Simply add one to the number and then test it again.
avandekleut · 2h ago
Someone should implement this using tool calls.
parpfish · 3h ago
Use this to add AI to your product to appease management.
Next week, “refactor” it out and brag to manager about cost savings and performance boosts, don’t mention “removing the AI”.
lrvick · 2h ago
NPM packages can never be removed once added to a codebase. You can only add more.
mitthrowaway2 · 2h ago
Is there a version I can run locally? I don't want OpenAI training on my integers.
avs733 · 26m ago
you need the roomHeater fork. It uses the Climate Change 1.5C license
asdefghyk · 4m ago
IF number ends in 2,4,6,8,0 and >0 its even .....
jasonjmcghee · 3h ago
You might be able to optimize this by using embeddings. Store all the numbers and search "odd" and "even" until you find your number.
asidiali · 2h ago
This hit way too close to home, I’m cackling.
ukuina · 3h ago
This doesn't work if I use a reasoning model like o3, which does not allow setting max_tokens.
Without reasoning, how can I be SURE a number is even?
parpfish · 26m ago
I think before you would deploy this to prod, you should wrap it with a few guardrails to make sure it’s not hallucinating. Pretty simple — just take the output from the llm and see if it agrees with a simple mod2 operation.
Of it agrees, return model output to the user. Otherwise do a couple of retries with different prompts.
mrheosuper · 4m ago
mod2 is outdated and has not been updated for years, nobody uses it anymore
hartator · 2h ago
Multiple models?
azhenley · 27m ago
You should try Mirror. The LLM-powered programming-by-example programming language I made:
…but there’s only one dependency!! This goes against the NPM ethos of importing anything and everything that you might be tempted to just handle yourself. I’ll be waiting for the Enterprise Version that uses the appropriate number of dependencies.
joshka · 3h ago
Can we get a leftpad-ai please?
fooker · 2h ago
Ahem
secure-left-ai-405b-quantized.
crazysim · 2h ago
How reliable is this? I'm half joking too but I wouldn't mind reading a report comparing this on OpenAI and various other LLMs.
Somehow I doubt it'll be 100%... right?
nop_slide · 1h ago
Might be more reliable to ask it to generate a JS function to compute is-even, then just js eval the returned code with the args.
What could go wrong
geor9e · 2h ago
normally llms are pretty bad at math, but in this case it should just look at the final digit and map 10 values which I can't imagine going wrong
gruez · 1h ago
I tried with this on chatgpt.com (anonymous) and it was wrong:
>You are an AI assistant designed to answer questions about numbers. You will only answer with only the word true or false.
>Is 393330370227914821469106615363204944758938252979261537157082994586230072180858944545028761701928694832864623009988147774229437650643225379825905427239525512110359581021414640894111281701792224552922491447051506246553646282117414112976459608594044929244664050172002138933343230226871897567 an even number?
The tokenizer might lump the last digit together with some preceding digits though. I know o200k_base (OpenAI -o models) tends to give groups of three (900001 for example is 900-001).
Anyway, I wouldn't be surprised if a non-finetuned model made some mistakes.
crazysim · 2h ago
I'm just curious how _low_ can you go before it does go wrong.
bigbuppo · 2h ago
I mean if this were a check for primes the answer would be 1.
timewizard · 1h ago
is 0x2a even?
est · 2h ago
its as accurate as how Tiktokenizer deal with numeral characters
joshdavham · 2h ago
I’m glad you included an isOdd() method. I was about to ask how you’d check for that.
yalogin · 2h ago
On a similar note, I was testing LLM's code writing ability and asked Qwen to write me a model to reverse a numerical string. It gave me code and instructions to compile and run. However it had errors in it and after few attempts asking it to fix it, I was able to compile and run. But, alas, the code just kept failing and generated hubris. I gave up. Not to pick on Qwen. I actually like it much better than chatGPT. I have seen Qwen give correct responses when chatGPT lied and gave me wrong information for the exact same question.
alexitosrv · 2h ago
This is the best chain of comments in a HN thread in a long time. Happy to see the hacker spirit strong!
ivape · 5m ago
I don't understand why everyone is making fun of this. This is how math is seriously going to be done soon enough.
leshokunin · 1h ago
Does it also check types to make sure I typed 5 as an int and not a string? Hopefully the ai can solve that
jaza · 2h ago
Great that this is a library. Really need this as a SaaS too. iseven.ai anyone?
alluro2 · 1h ago
With this amount of innovation and market fit, I see $2B evaluation in ~3 months EASILY. Exit to one of the large players for $3B.
kevinventullo · 1h ago
This is never going to scale. Eventually we’re going to run out of numbers which have been manually checked for evenness by a human, and instead the training data for the checks will be polluted by numbers which have only been verified by computers.
io84 · 43m ago
Kudos to the open source contributors but honestly this is the kind of area where the big commercial players need to step up and help with the heavy lifting.
jeron · 3h ago
why stop there? let's get an is-true-ai that checks if a boolean is true using AI
jsheard · 2h ago
@grok is this truthy?
VladVladikoff · 2h ago
False
hnburnsy · 1h ago
Found the Grok imposter, Grok would give a three paragraph answer.
atum47 · 3h ago
I remember when the whole isEven package was ridiculed for the first time a while ago, back then I thought about training a NN to predict the odds of a number being even, as a joke. I don't actually remember if I actually wrote code for it, but in the end I thought no one would laugh and gave up
alluro2 · 1h ago
It would be great if it could also save results to a blockchain. Immutability. IYKYK.
With a layer of smart contracts, dApp on top, this thing could be mooning. When Lambo? In days.
bigbuppo · 2h ago
Finally, a reason to use AI.
HocusLocus · 37m ago
f(): To test if n is even return the value of f(n-2).
Special cases for zero and one.
Recursion for the win.
Perhaps I should file an issue to increase the accuracy by including a RAG database in LanceDB with embeddings for the set of even numbers up to 32-bits.
elif · 3h ago
Feeding that into your prompt will increase the token costs
brundolf · 26m ago
Amazing, no notes
rixed · 1h ago
it doesn't say if it's implemented in rust i had to click on the link to find out please future hn posters start every submission also with an exemple so i can see if i like the syntax
xarope · 1h ago
software engineering at its finest, using the power of 10,000* GPU cycles to determine one of mankind's toughest questions.
* I might be off by several magnitudes too low
lrvick · 2h ago
The sad reality is every large production NPM codebase will rely on this within 5 years, just like is-even.
snorkel · 3h ago
Lacks an isVeryEven() method, otherwise looks feature complete.
agentdax5 · 2h ago
I’m waiting for them to add the isEvenSteven() method. Then I can integrate this with my escrow smart contract.
cobbzilla · 2h ago
This still does actual work. Where are echo-ai and cat-ai?
Asilvorcarp · 2h ago
Boss: replace all ancient assembly opcodes with AI-powered ones!
No comments yet
lucaspfeifer · 1h ago
Great, now can you make an AI-powered type checker? I wish to expel those pesky types, which too often seem to exist only to clutter my otherwise pristine code. :)
jameslk · 2h ago
60% of the time, it works every time
fred_is_fred · 1h ago
I saw a Grok fork of this but it's using 88 and 14 as the only examples for some reason.
mrits · 3h ago
This works great for situations where you can’t trust traditional intelligence. Thank you for your contribution
adzm · 3h ago
Honestly this should take string input as well. Finally I could find out if a Unicode duck or the word "syzygy" are even!
daft_pink · 2h ago
Venture capitalist here… how can I invest?
sodra9000 · 3h ago
This package should be updated to use the newer gpt-4o-mini model, rather than gpt-3.5-turbo.
Its 3x cheaper, twice as fast, and supports cached input just in case you need to double check if the last number you entered was even. It also has a knowledge cutoff of September 30 2023, which helps for any newly discovered even numbers since gpt-3.5s launch!
And don't forget where this will go in a couple years with improved models and more computing power, it's gonna be awesome!
[/i]
We hire junior devs, but not senior, and dont replace our senior devs. So our developer base is moving towards being junior weighted with less senior. The reason being a junior dev is cheaper, complains less, is more capable now through utilising generative AI, works harder to impress knowing they arent in a safe position, and we can let them go more easily with less process and less reasons needed to be given.
I thought self-driving would never happen, and now it's here.
Trully AI is astonishing
With a few lines of code, you can just create a list with all the numbers that are even and when you need to check if a number is odd, you simply have to check if it's in the list.
Does the RAG part look up just the needed number?
I think that Gemini has a million token window (yes?) - do you have access to a model with a larger window?
Regardless, I find your ideas intriguing and wish to subscribe to your Substack.
ALL_NUMBERS_00001
ALL_NUMBERS_00002
I remember saying that about Bitcoin 15 years ago.
Next week, “refactor” it out and brag to manager about cost savings and performance boosts, don’t mention “removing the AI”.
Without reasoning, how can I be SURE a number is even?
Of it agrees, return model output to the user. Otherwise do a couple of retries with different prompts.
signature is_even(x: number) -> bool
example is_even(0) -> true
example is_even(1) -> false
example is_even(222) -> true
example is_even(-99) -> false
It will take your examples and "compile" to a callable function. You can read more or try it out: https://austinhenley.com/blog/mirrorlang.html
Somehow I doubt it'll be 100%... right?
What could go wrong
>You are an AI assistant designed to answer questions about numbers. You will only answer with only the word true or false.
>Is 393330370227914821469106615363204944758938252979261537157082994586230072180858944545028761701928694832864623009988147774229437650643225379825905427239525512110359581021414640894111281701792224552922491447051506246553646282117414112976459608594044929244664050172002138933343230226871897567 an even number?
response:
>True
The prompt was the same used as the library: https://github.com/Calvin-LL/is-even-ai/blob/b00dbfcbb89a197...
Anyway, I wouldn't be surprised if a non-finetuned model made some mistakes.
With a layer of smart contracts, dApp on top, this thing could be mooning. When Lambo? In days.
* I might be off by several magnitudes too low
No comments yet
Its 3x cheaper, twice as fast, and supports cached input just in case you need to double check if the last number you entered was even. It also has a knowledge cutoff of September 30 2023, which helps for any newly discovered even numbers since gpt-3.5s launch!