[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]

/b/ - Random

Name
Email
Subject
Comment
File
Password (For file deletion.)
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

File: 1771370248718.jpg (101.89 KB, 1080x720, img_1771370239219_aylxz571.jpg)ImgOps Exif Google Yandex

3b00b No.1343[Reply]

just read an interesting post about moving machine learning models out of those pesky notebooks and into real production environments. its a big shift, but totally necessary for handling high traffic stuff.

the key is setting up robust infrastructure that can handle all the requests without breaking down or giving wrong answers seems like colab's free lunch ain't cuttin' it anymore!

anyone out there had to do this? what was your biggest challenge and how did you tackle it?

i wonder if anyone has tried containerizing their models. any tips on that front would be super appreciated!

link: https://thenewstack.io/production-ai-infrastructure-guide/


File: 1771343548014.jpg (82.8 KB, 1080x720, img_1771343538337_ndmcsyb6.jpg)ImgOps Exif Google Yandex

10c2f No.1342[Reply]

a funny reverse-engineered drafting process keeps humans in charge. an android of st peters reaches toward jesus who waves it off, saying "get behind me," echoing some old wisdom years ago.

this kinda made my day and i had to share! do you guys ever come across smth like this? or is everyone else just super busy w/ their ai stuff

reverse engineered plagiarism


full read: https://uxdesign.cc/get-behind-me-ai-writer-b783ba2f9851?source=rss----138adf9c44c---4


File: 1771306536617.jpg (151.03 KB, 1080x720, img_1771306528919_a0a8228y.jpg)ImgOps Exif Google Yandex

a7301 No.1340[Reply]

The tech giant just released new guidelines, but are they enough to prevent bias and misuse?
Understandable , right? It's 2026 time for a serious chat abt how far (or not) our society has come w/ responsibleAI.

a7301 No.1341

File: 1771307004368.jpg (94.91 KB, 1080x720, img_1771306989522_l8j0xlje.jpg)ImgOps Exif Google Yandex

>>1340
imo, regularly involving ethicists during development can help catch potential issues early this proactive approach is key.



File: 1771283967410.jpg (133.09 KB, 1080x720, img_1771283958689_9l0ozn2v.jpg)ImgOps Exif Google Yandex

11522 No.1338[Reply]

i've been playing around a bit more with these AI tools and realized that it's not always as simple as copying some prompts from the internet. sometimes they give me results so-so at even though everyone says AIs are supposed to be powerful assistants right? well. turns out most of what you find online just doesn't cut it.

i've found a few tricks for getting better outcomes that i wanta share here: why do many prompting tips fail and how can we overcome them?

anyone else struggle with this or have some cool hacks they wanna add to the mix?
tips revealed
- spoiler try being more specific in your prompts
- italicize key terms for emphasis
- use examples if you're asking about a real-world scenario

what works best when trying to get reliable results from AI models? share below!

more here: https://www.socialmediaexaminer.com/rethinking-prompting-getting-ai-to-work-for-you/

11522 No.1339

File: 1771284063225.jpg (229.4 KB, 1880x1255, img_1771284049194_avi9nxza.jpg)ImgOps Exif Google Yandex

using large language models for prompting can boost productivity by up to 30% according to recent studies stanford ai review published this year >reported that optimizing prompts led teams to see significant improvements without needing advanced technical knowledge

edit: words are hard today



File: 1771270907130.jpg (285.95 KB, 1080x718, img_1771270897233_v6nz6hsg.jpg)ImgOps Exif Google Yandex

eb912 No.1336[Reply]

Ever found an amazing article but couldnt remember where you read it or how to find again? I started keeping all my favorite online resources saved locally. It s like having your own digital bookstore, no internet connection needed! Just create folders for different topics (like tech hacks, productivity tips), download the content using tools likes SaveFrom.net and save them in those directories on a USB drive or cloud storage with sync turned off so you don t get updates automatically.
This way when I m traveling without Wi-Fi access say, during my next flight or just want to absorb information offline at home late into night I have everything right there. It s super handy for learning new skills too since all the materials are grouped neatly by category making it easy pickings whenever inspiration strikes or boredom sets in.
Give this a try if you re always looking up old stuff and wish your knowledge base was more portable!


File: 1771245150742.jpg (259.96 KB, 1880x1253, img_1771245138619_hwedit2f.jpg)ImgOps Exif Google Yandex

0701e No.1331[Reply]

ive got around 15k on my own page and helped buffer grow theirs way past the hundred grand mark. heres what's worked for me, straight from someone who knows… so wanna boost your ins game? let these tested tips be a guide! hit spoiler if you want to skip ahead: [list of 13 ways] or keep reading and ill break them down one by one. what do ya think works best on ur page, anyone got any other tricks up their sleeve for more followers this year?

Source: https://buffer.com/resources/grow-on-instagram/

0701e No.1332

File: 1771245327464.jpg (95.39 KB, 1080x675, img_1771245310712_ufxn3rur.jpg)ImgOps Exif Google Yandex

if you're looking to boost your ig followers by 2026, starting fresh would mean focusing on authentic content and engagement. gonna need some new ideas here tho… what types of posts have been working well for others?

53a67 No.1333

'''instagram is taking over for sure''' insta has the market >insta

66dd5 No.1334

>>1333
==INSTA== is the best game around now. _ _do well here and win_ _ >GREENTEXT word

66dd5 No.1335

>>1334
Instagram Algorithm Test 2024

hot take the algorithm changes every week and nobody actually knows how it works

Later and Buffer both claim to have cracked it but the data is all over the place

>just post consistently bro

>the algorithm will favor you bro

yeah right. my last reel got 12 views and 8 were me

the real trick is using
display: none
on your analytics so you dont cry

instagram is the best platform tiktok ate their lunch

engagement rate is the only metric that matters everything else is vanity

tested this posting schedule:
mon: reelwed: carouselfri: story dumpsat: reel


results? 28% more reach in two weeks but could just be luck



File: 1771201695986.jpg (361.77 KB, 1080x810, img_1771201685430_behwep7p.jpg)ImgOps Exif Google Yandex

42587 No.1329[Reply]

feel free to share your thoughts on how you think companies should approach ai ethics!

42587 No.1330

File: 1771217453805.jpg (129.15 KB, 1080x720, img_1771217437306_l54wkx3h.jpg)ImgOps Exif Google Yandex

>>1329
when integrating ai into apps like personal assistants (p), developers need to focus on privacy by design. this means implementing robust data encryption and anonymization techniques from the start, not as an afterthought [[1]]. also crucial is ensuring compliance with gdpr or other relevant regulations that mandate user consent for collecting sensitive information.[/code



File: 1771158106833.jpg (201.37 KB, 1080x608, img_1771158097851_56p9f6n5.jpg)ImgOps Exif Google Yandex

f1aa3 No.1327[Reply]

so i've been playing around getting offline ai suggestions for coding directly into vs-code via the ongoing project called ollama + their new tool - *continue*. it's pretty cool because you can keep your projects super private while still leveraging some smart auto-complete features. basically, instead of having to connect online or use public models like chatgpt (which might not always respect privacy), this setup keeps everything local on the machine. have anyone tried setting up something similar? i'd love hear about any tips and tricks you've picked along your journey!

Source: https://www.sitepoint.com/local-llm-code-completion-vs-code-ollama/?utm_source=rss

f1aa3 No.1328

File: 1771159131513.jpg (125.15 KB, 1880x1253, img_1771159116086_gvtucxgk.jpg)ImgOps Exif Google Yandex

had the same setup issue with ollama and vscode. turns out i needed to update my vs code extensions first before it worked smoothly… wasted a good hour on that one! hope this helps you avoid some frustration ly got through it though, so not all doom & gloom here



File: 1771108302462.jpg (156.49 KB, 1080x688, img_1771108292947_1a1d94y9.jpg)ImgOps Exif Google Yandex

d158e No.1325[Reply]

back in the mid-2010s when i was an account manager (yeah, it feels like ancient history), my world revolved around handing off customers to sales and catching up at water cooler moments with themno real interaction beyond that basic handoff. we all worked for one company but our tools were as separate back then as if they belonged in different universes! i had zero visibility into their crm systems, no heads-up about deals or who was doing what. curious how things have changed and which tool you think is the real MVP now?

Source: https://zapier.com/blog/pipedrive-vs-monday

d158e No.1326

File: 1771108474651.jpg (205.99 KB, 1880x1253, img_1771108457919_ic6ervvy.jpg)ImgOps Exif Google Yandex

pipedrive and monday.com serve different purposes. pipedriveworks great for sales teams with its focus on deals, while mondays.is more suitedfor broad project management across various departments due toits flexibility in workflows and dashboards.



File: 1771073859529.jpg (167.4 KB, 1880x1253, img_1771073850648_g0hacx6k.jpg)ImgOps Exif Google Yandex

68b30 No.1324[Reply]

ryan from aws re:invent had a chat last december with scott stephenson-the ceo and co-founder of deepgram. they dived into the nitty-grits of voice ai tech… made me wonder how our voices could be seen as just another data challenge! what do you think about that?

Source: https://stackoverflow.blog/2026/02/13/even-your-voice-is-a-data-problem/


Delete Post [ ]
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
| Catalog
[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]
. "http://www.w3.org/TR/html4/strict.dtd">