HNHacker News
TopNewBestAskShowJobs

seanthemon

717 karma · joined May 3, 2016

I like rocks
submissionscomments
seanthemon··on Show HN: Android Hacker News Widget and App
I can't seem to find it on the app store, just shows not found
seanthemon··on FindMyCat – Open-Source Pet Tracker
Could your cat perhaps have heard the sound of the whine of the electronic components? Maybe an interference with your phone and other electronics close by. Just a shot in the dark
seanthemon··on RestGPT
I feel like that script needs a few 'are you completely sure?' Prompts
seanthemon··on Sucking carbon dioxide out of the sky is moving from science fiction to reality
So.. barbeque time, right?
seanthemon··on 'How do you know when your art is good?'
This is the way, deep study of old and new artists you admire, and looking through lenses
seanthemon··on Show HN: Host a Website in the URL
Of course! Made in Britain
seanthemon··on The endless battle to banish the world’s most notorious stalker website
https://archive.ph/HJaaz
seanthemon··on WhatsApp’s founder first pitch in the community
Shows the importance of MVP and going with the market flow - interesting to read the comments and see how it went under the radar for so long before becoming a worldwide necessity in some cases
seanthemon··on A complete guide to fine-tuning Code Llama
This is a great article, but wondering how these datasets are made - I would love to make a fine-tuned model on our own code base
seanthemon··on Ask HN: What is your policy regarding smartphones for your children?
Agreed, I was born in the 90s and I spent 90% of my time glued to my pc, tv, psp, gameboy, nokia...
seanthemon··on Keeping Figma Fast: perf-testing the WASM editor
Figma is beyond chrome, it's a quantum app: it's both in a state that will lock up my pc while everyone on zoom waits for the screen to load and a journey in following designer's instructions: "no a bit more left, no actually right, ok go down, no up, ok back go the left a bit, ok here it is!", same vibe as rally racing lol
seanthemon··on Giving up the iPad-only travel dream
How's the cost? I'm thinking of running a docker instance from my pc and connecting to it remotely
seanthemon··on Code Llama, a state-of-the-art large language model for coding
There's also code golf stackexchange[0] and iirc reddit has some daily programmer challenges too

[0] http://codegolf.stackexchange.com/

seanthemon··on Giving up the iPad-only travel dream
Which cloud provider do you use? I'm currently connecting to my pc but i don't like this solution. I'd like to run a docker instance
seanthemon··on AI isn’t good enough
Very insightful point - do you think that OpenAI and Microsoft are aiming at developers so that we may spill this knowledge by N-shotting the chat query? I.e if the AI doesn't understand some fundamental knowledge, often i'll teach it so that it can be more effective at helping me, inadvertently I spill some of this tucked knowledge.
seanthemon··on Show HN: Shimmer – ADHD coaching for adults, now on web
Or highest paid best coach..
seanthemon··on AI isn’t good enough
Curious, how does it do when you give it examples / use playground.openai?

I find that no examples often leads to the same result as a befuddled junior, but with examples often it gains confidence. Also I find playground to give me much better code snippets than chat sometimes

seanthemon··on Threads.net Can Go to Hell
You have been cssuspended
seanthemon··on Odd and unsafe baby car seats from the past
One strand away from becoming Voaldahlmort
seanthemon··on We’re all just temporarily abled
Thank you, i'm so glad your MIL made such a wonderful recovery!

We're about a year in and he's gotten to the "I can do this myself, I'm not depressed" phase, haha!

seanthemon··on We’re all just temporarily abled
My dad went from someone I can chat with every day about something or build something together to partially disabled due to first stroke and now unable to communicate fully or easily due to the second stroke at 65.

He lost his job and sadly there are few things to help with those who are unable to communicate.

This all happened over a single weekend.

seanthemon··on Show HN: AI transforms eating – healthier choices in your busy life
I'm going a little critical in this comment, i'm hoping to give my first insights and some improvements I would like to see without scaring you off.

I'm confused/put off for a few reasons:

1. I can use AI to generally create a meal plan for my whole family if needed, why is this better? Esp if paid version only gives me 2 diets, so if I have kids, gotta pay up?

1.1 I can share these diets with a dietician, setup a consult for an hour and pay about 3 mo of your first level plan to get a fully approved diet from someone in the diet industry, and often these dieticians want to see YOU make your OWN diets to fit your lifestyle within their guidelines

2. Your website states that it's people making my diets and not AI, who are these people? Are they registered dieticians? Can we see their credentials? (I got this scepticism from 'our team will' and the manual diets options) Also in what ways are AI generating my diets?

3. Your diet form is extremely rudementary, it doesn't cover what people are currently eating, substitutions, medical blockers (diabetics, high cholesterol, low sodium diets)

4. Your website gives no example diets, just one video where you get a very generic looking diet

If I were to use this package I need to see a few things: 1. Registered dietitians working or consulting for your company, their credentials and faces on the product

2. A much more in depth form, not just a focus on weight loss for a normal person, it needs to cover a huge range of cases (dieticians can help with this)

3. Don't put number of generated diets under a paid feature, max it out at 3, family plan 8 and unlimited if you want - this is important because dieting is evolution, not revolution.

4. History, as my diet changes over time and my conversations with dieticians over my concerns

5 (extra). Private log book, reminders, grocery list generation, advanced substitutions (change specifically some ingredient, some snack), eating windows (for fasting, Ramadan, etc), holiday special diets to help you navigate the difficult seasons, etc

Thank you for sharing.

seanthemon··on CloudNordic Status – Ransomware Attack
That's a hurtful thing to read, something that can kill a company.

I'm a cloud junker, always been on AWS or similar, do any older gurus know what the right approach would've been? Check every drive? Separate them into sandboxes?

seanthemon··on The rise of the open kitchen
History: greeks and romans traditionally had separated kitchens from homes due yo the use of hearths. Another aspect was fire. Kitchens used wattle chimneys which were super flammable, so it was placed as far away from the main home as possible - also kitchens were a lot noisier.
seanthemon··on Show HN: Create your own Discover Weekly
I like the concept, but I feel a bit scared to pop in my credentials - I see no reason to trust this site. Ideally i'd love to give a list of songs I have and for your app to generate the discover weekly so I can build trust.
seanthemon··on Cyberattack shutters major NSF-funded telescopes for more than 2 weeks
Scientists (and oft most users..) don't want friction when doing work, so they find ways around security protocols to speed up their interactions leaving little security holes everywhere.
seanthemon··on Artists have forgotten how to draw
Honestly this is more the difference between putting your personality to your art vs just your art - I also post my paintings online but I hardly get the views of people making shorts with their face and revealing the artwork, usually no matter how mediocre it will have a lot of followers.
seanthemon··on Show HN: LlamaGPT – Self-hosted, offline, private AI chatbot, powered by Llama 2
Many ways to skin this question, but in essence a simple idea is that word vectorization is assigning a numerical representation to a specific word, embeddings on the other hand are taking those words, turning them into numerical representations but keeping semantically similar words closer dimensionally.

Yes, turning words into vectors is it's own class of machine learning. You can learn a lot on the NLP course on hugging face https://huggingface.co/learn/nlp-course/chapter1/1 (and on youtube).

seanthemon··on Testing on production
It's so closely related and so often mistaken, feature flags and tenant features are two completely different things
seanthemon··on Far Manager: files and archives in Windows
real recognizes real
← PreviousPage 7 of 10Next →