Joined November 2016
- Tweets 8,189
- Following 693
- Followers 3,931
- Likes 20,248
479 Photos and videos
Messed up my RPI so bad it won't boot up via SD card nor SSD. Idea was to upgrade the setup with Pironman 5 and it completely messed my existing setup. So disappointed! ๐
56
Prateek Gogia ๐ญ retweeted
May 25
Spec driven development feels good in the same way that writing a todo list feels good (before reality hits)
1
3
33
2,342
Prateek Gogia ๐ญ retweeted
ssh was created in 1995
tmux was created in 2007
sadly most engineers won't understand why this is relevant
179
554
13,555
2,090,865
Prateek Gogia ๐ญ retweeted
May 22
Software Devs making a comeback at Starbucks like
May 21
JUST IN: Starbucks retires AI inventory tool across North America after it reportedly miscounted & mislabeled store items.
47
583
10,645
1,430,889
I have had done something similar what Neha suggests here and lately I have been doing a lot of self hosting stuff. I started with buying a Pi 5 and later on added other components and learned more stuff around what is used for what.
May 16
I started my career as web and graphic designer and then moved to front-end. 5 years ago, I decided to move out of front-end or basically - upskill and I started Cloud.
I did enrolled in course for cloud and I wasted my money. As someone who was coming from front-end background, it was hard to keep up with the folks in course who are already pro in backend, devops etc.
So, I started doing self learning & it helped me a LOT.
If youโre a beginner and want to get started with cloud platforms, architecture design, or system design, start with small hands-on projects.
I will share 1 project idea for next few days (it would be a note for my my blog too)
PS: I will be using AWS. You can do same in any cloud platform.
๐งต
1
4
358
While it can be challenging to keep up to date with what's hot in the stack because of AI revolution I also feel that it has opened doors for people like me to just explore territories which were completely unknown few years back.
1
1
65
Sometimes I messed up real bad when I was trying networking stuff during adguard or pihole integration because my wifi stopped working while me messing with DHCP and I kinda panicked because of that but claude was a huge help that time.
1
149
Prateek Gogia ๐ญ retweeted
May 12
how Iโm getting work done without freaking out about the npm supply chain today:
โฏ go run hello_world.go
hello world
7
4
167
7,175
We're winding back our peak hours limit reduction and doubling 5 hour limits.
Excited to partner with SpaceX to bring you more compute and we'll keep pushing to bring you the best coding agent in the world.
Weโve agreed to a partnership with @SpaceX that will substantially increase our compute capacity.
This, along with our other recent compute deals, means that weโve been able to increase our usage limits for Claude Code and the Claude API.
358
122
3,867
244,982
Prateek Gogia ๐ญ retweeted
May 5
anyone else getting this issue with their AI assistants??
135
222
8,664
357,853
If youโre in Berlin or nearby, you can make the most of your Deutschlandticket and hop on a RE or similar train to visit the beautiful city of Schwerin in northern Germany.
1
170
Hmm..so all this fear in the market for last 2 years just for nothing? ๐คก
May 1
we want to build tools to augment and elevate people, not entities to replace them.
94
Prateek Gogia ๐ญ retweeted
Apr 28
when the API says no, the DOM says yes
4
7
204
44,440
Prateek Gogia ๐ญ retweeted
Apr 27
Its the beginning of the end of subsidized AI subscriptions. GH Copilot is moving to usage-based billing, as has Claude (for business customers.) Fair to assume more will follow.
I expect this change will also be a great boost for open models - cheaper, and pretty good already
Starting June 1st, GitHub Copilot will move to a usage-based billing model as GitHub Copilot supports more agentic and advanced workflows.
In early May, you'll see a preview bill experience, giving visibility into projected costs before the transition.
๐ Read more about the upcoming change: github.blog/news-insights/coโฆ
127
281
4,249
511,949
Outsourcing static content generation to a service like groq was a good choice. Running LLMs locally is expensive and painful.
1
73
Being an absolute noob, I tried running some open source models using ollama in my pi5. Probably a stupid idea to get the most of a machine of that size but a good idea to understand how LLMs are working under the hood.
1
146
also explains why everyone on this planet is complaining about RAM prices. I joined the same league last night.
63