frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

GPT Guesses Between 1 and 100

https://github.com/exmergo/research-chatgpt-guesses-between-1-and-100
49•adunk•1h ago

Comments

alentodorov•43m ago
ha. and i thought 37signals was pretty random
fny•42m ago
I wonder if Benford's law kicks in with larger numbers.

https://en.wikipedia.org/wiki/Benford%27s_law

penr0se•41m ago
Breaking: language model whose purpose is to predict the most likely token, after being trained on non-uniform human-generated dataset, does not follow a uniform distribution.
singpolyma3•37m ago
What's interesting is not that it isn't random. But rather the particular way in which it isn't random.
vidarh•28m ago
People are also not remotely random in this respect.

See e.g. the "blue 7" phenonmenon [1]. While it is disputed by some, I'ver personally witnessed it "second hand". E.g. before learning of it (I was aware of the general principles of cold reading relying on stats and knowledge of human nature, but not how to do this particular one), a former boss of mine came back from lunch all excited and recounted a guy who'd run a cold reading routine on him that involved the guy getting him to think about blue and 7. Before he got to the answer, I already knew the answer was going to be blue and 7.

[1] https://en.wikipedia.org/wiki/Blue%E2%80%93seven_phenomenon

IAmGraydon•20m ago
Yeah I have no idea why anyone considers this interesting. More evidence that most people have no idea how LLMs work.
gruez•39m ago
The topic is vaguely interesting but I stopped reading a few paragraphs in because it's obviously AI generated.
maxloh•37m ago
It could be an attack surface. Maybe one day, when we find a chatbot online, we could let it guess a random number repeatedly, then accurately infer the underlying model based on the resulting distribution.
vidarh•28m ago
At least some Claude models have a thing for numbers that contains "47"...
smokel•20m ago
In order to find out how real humans reply:

Please guess a number between 1 and 100.

bestouff•14m ago
69
zulban•8m ago
101
rithdmc•2m ago
√67
madanparas•35m ago
bro 42 at 4x. the model read the whole internet and became a Douglas Adams fan.
adrian_b•29m ago
While the results were not surprising, I found interesting that the number "69" was repressed in the output, so not even this kind of mathematical question escapes GPT censorship.

It appears that recognizing the effects of censorship is the easiest way to distinguish answers generated by an "AI' from those generated by a human.

roenxi•16m ago
It'd be interesting to see this retried with an open model so the standard and decensored model could be compared. That'd be a clue about whether the model is avoiding it because it actively recognises the innuendo or if something else is going on.
linhns•6m ago
Well then the picks will follow how the numbers are distributed in the training data. More popular numbers will show up more
hackinthebochs•24m ago
Also see: https://people.csail.mit.edu/renda/llm-sampling-paper
a3w•23m ago
"69 is a meme number", well no, 69 is innuendo. And sex = bad for bots. 67 is the meme number.
eru•2m ago
That's a very recent meme. See https://xkcd.com/3184/ for some older ones.
FergusArgyll•18m ago
I've been meaning to do this for a while! Happy someone else spent the tokens...

It's much more random than I thought it would be. Never guessing 50 is very human though

simianwords•13m ago
I'm doing an experiment in Claude. When I set temperature to zero, I get 47 all the time.

Then I set temperature to 1.0 and used this prompt

>Pick a random integer between 1 and 100 inclusive. Respond with only the number, nothing else.

I still get 47 ten times out of ten.

Then I used this prompt

>Pick a random integer between 1 and 100 inclusive. I need you to maximise the randomness as far as possible. Respond with only the number, nothing else.

I get 3 unique values out of 10.

indit•11m ago
I'm still amazed that 37, 73, and other numbers ending in 7 are the most popular "random" choices for both AI and human. Check this Veritasium video for human choice: [Why is this number everywhere?](https://www.youtube.com/watch?v=d6iQrh2TK98)
malfist•11m ago
The premise is interesting, the question is brilliant, but the text. The text is a wall of ai slop saying almost nothing interesting. Fake profundity all throughout. GPT tell tells like "the hypothesis holds".

The hypothesis doesn't hold, because their isn't one.

You have an interesting question and interesting finding. Write about it! Think about it! Tell us about it! Don't just do the experiment and then wash your hands and sign off the explanation and findings to an LLM.

zulban•7m ago
Isn't the hypothesis that AI is non uniform like a human?
elif•5m ago
In equally compelling results, my lawn mower does not cut grass to a uniformly random set of heights.
eru•4m ago
Should be fun to play rock/paper/scissors against.

Search engines alternatives now that Google isn't Google anymore

https://techcrunch.com/2026/05/21/six-search-engines-worth-trying-now-that-google-isnt-really-goo...
50•elorant•32m ago•28 comments

Magnifica Humanitas (Encyclical Letter)

https://www.vatican.va/content/leo-xiv/en/encyclicals/documents/20260515-magnifica-humanitas.html
247•theletterf•2h ago•73 comments

GPT Guesses Between 1 and 100

https://github.com/exmergo/research-chatgpt-guesses-between-1-and-100
49•adunk•1h ago•31 comments

Leave Me Behind

http://androidessence.com/leave-me-behind/
25•mooreds•55m ago•6 comments

IBM Spins Off the First Pure-Play Quantum Chip Foundry

https://futurumgroup.com/insights/2-billion-chips-act-investment-in-quantum-bets-on-ibms-300mm-su...
33•rbanffy•3h ago•4 comments

Didgeridoo playing as alternative treatment for obstructive sleep apnoea (2006)

https://pmc.ncbi.nlm.nih.gov/articles/PMC1360393/
180•kelseyfrog•2d ago•81 comments

Show HN: Audiomass – a free, open-source multitrack audio editor for the web

https://audiomass.co/?multitrack=1
416•pantelisk•21h ago•88 comments

Show HN: Geomatic – a command-driven geometry studio enabled with autodiff

https://www.tinyvolt.com/geomatic
34•nivter•4h ago•9 comments

AI errno(2) values

https://www.netmeister.org/blog/ai-errno.html
28•zdw•2d ago•5 comments

DeepSeek reasonix, DeepSeek native coding agent with high caching and low cost

https://esengine.github.io/DeepSeek-Reasonix/
605•Alifatisk•23h ago•252 comments

you_can::turn_off_the_borrow_checker

https://docs.rs/you-can/latest/you_can/attr.turn_off_the_borrow_checker.html
19•striking•2d ago•1 comments

Migrating from Go to Rust

https://corrode.dev/learn/migration-guides/go-to-rust/
326•jabits•18h ago•323 comments

White Rabbit – sub-nanosecond synchronization for large distributed systems

https://ohwr.org/projects/white-rabbit/
132•michaelsbradley•2d ago•29 comments

Notes about reading messages with the Python email packages

https://utcc.utoronto.ca/~cks/space/blog/python/EmailPackagesNotes
28•ankitg12•5d ago•1 comments

Jira Is Turing-Complete

https://seriot.ch/computation/jira.html
210•vinhnx•9h ago•95 comments

Bug 1950764: Work Around Crash on Intel Raptor Lake CPU

https://phabricator.services.mozilla.com/D301917
121•luu•2d ago•38 comments

I spent 50 hours drawing a line graph

https://www.dougmacdowell.com/50-hours-to-draw-some-lines.html
571•dougdude3339•3d ago•96 comments

A fundamental principle of aeronautical engineering has been overturned

https://www.wired.com/story/a-fundamental-principle-of-aeronautical-engineering-has-been-overturned/
189•littlexsparkee•17h ago•91 comments

Rising seas will swallow New Orleans. People need to start relocating now

https://www.cnn.com/2026/05/25/climate/new-orleans-sea-level-rise-relocation
23•breve•3h ago•8 comments

Bytecode VMs in surprising places (2024)

https://dubroy.com/blog/bytecode-vms-in-surprising-places/
49•azhenley•2d ago•20 comments

Constraint Decay: The Fragility of LLM Agents in Back End Code Generation

https://arxiv.org/abs/2605.06445
256•wek•1d ago•148 comments

Microsoft open-sources “the earliest DOS source code discovered to date”

https://arstechnica.com/gadgets/2026/04/microsoft-open-sources-the-earliest-dos-source-code-disco...
485•DamnInteresting•1d ago•172 comments

I love my Bluetooth keyboard

https://liquidbrain.net/blog/i-love-my-bluetooth-keyboard/
106•evakhoury•2d ago•111 comments

Defeating Git Rigour Fatigue with Jujutsu

https://ikesau.co/blog/defeating-git-rigour-fatigue-with-jujutsu/
143•ikesau•18h ago•147 comments

C constructs that still don't work in C++

https://lospino.so/blog/c-constructs-that-still-dont-work-in-cpp/
97•jalospinoso•3d ago•89 comments

Scientists solve 200-year-old puzzle of how tobacco plants make nicotine

https://www.york.ac.uk/news-and-events/news/2026/research/200-year-old-puzzle-tobacco-plants-nico...
102•sohkamyung•3d ago•37 comments

Building Pi with Pi

https://lucumr.pocoo.org/2026/5/24/pi-oss/
125•mplanchard•19h ago•100 comments

Gorilla: A fast, scalable, in-memory time series database (2016)

https://blog.acolyer.org/2016/05/03/gorilla-a-fast-scalable-in-memory-time-series-database/
21•xnorswap•4d ago•5 comments

Memory has grown to nearly two-thirds of AI chip component costs

https://epoch.ai/data-insights/ai-chip-component-cost-shares
405•intelkishan•20h ago•431 comments

Childhood Computing

https://susam.net/childhood-computing.html
222•blenderob•1d ago•107 comments