PS Chua
banner
pschua.bsky.social
PS Chua
@pschua.bsky.social
#B2B marketer
#Python programmer
#Digitalmarketing professional ( #GoogleAds #LinkedIn #Meta )
Part time gamer
pschua.com
"We live in a metrics obsessed culture that is obsessed with narrowly defined productivity. There's too much focus on accountability and too little focus on creativity."

#innovation #research
Why Bell Labs Worked.
Or, how MBA culture killed Bell Labs
open.substack.com
May 18, 2025 at 3:52 AM
"The 'old world thinking' that gave us the original horseless carriage was swapping a horse out for an engine without redesigning the vehicle to handle higher speeds. What is the old world thinking constraining these AI apps?"
#ai #llm
AI Horseless Carriages | koomen.dev
An essay about bad AI app design
koomen.dev
May 1, 2025 at 3:07 AM
The feeling when shipping a side project can't be replicated elsewhere.

"I think we exist to bring new things into existence. If you ask me, to the extent there is a meaning of life, that’s it. We exist to create."

#development #sideproject
The blissful zen of a good side project
One of life's greatest simple pleasures is creating something just for yourself.
joshcollinsworth.com
April 13, 2025 at 12:43 PM
The higher context window is certainly a big win. It's frustrating the number of times Claude warns me that I'm reaching the limit. I noticed that Gemini retains its answering ability deep into the context window, while ChatGPT and Claude seems to degrade.
#llm #Gemini #Claude
Gemini 2.5 Pro vs. Claude 3.7 Sonnet: Coding Comparison
This blog-post compares the coding capabilites of new Gemini 2.5 Pro experimental and Claude 3.7 Sonnet (thinking)
composio.dev
April 4, 2025 at 2:24 PM
Excellent piece by @birgitta410.bsky.social about the common missteps of LLMs for coding. The write-up does a great job of capturing most of what others and I have been through.

"AI sometimes took brute-force approaches to solve issues rather than diagnosing what actually caused them."
NEW POST

To work effectively with agentic coding assistants, Birgitta Böckeler found she needs to intervene, correct and steer all the time. She describes examples of these interventions indicating the skills we need to correct the tools' missteps

martinfowler.com/articles/exp...
March 31, 2025 at 8:10 AM
Hmmm... I wonder which of these will create an image successfully.

#openai #gpt4o #imagegen #imagegeneration #llm
March 27, 2025 at 11:18 AM
Reposted by PS Chua
This is, btw, Microsoft retreating from OpenAI.
TD Cowen reports that Microsoft has actually walked away from TWO GIGAWATTS of compute capacity, owing it to them deciding not to support OpenAI's training workloads, and Microsoft being in an "oversupply position." It's even worse than I previously thought!
March 26, 2025 at 2:37 PM
"All in all, this leads to an important point that people are overlooking- so much of AI Engineering isn’t an intelligence play; it’s an architecture play. Even many of the older models have a lot of untapped intelligence that can be unlocked by good design."

#ai #llm #agents
AI Market Intelligence: March 2025
What are the most important AI Developments to watch for Founders and Investors
open.substack.com
March 27, 2025 at 1:08 AM
Another great use case for the latest image gen capability of gpt-4o
One of the best use cases for OpenAI ChatGPT-4o’s image generation feature is creating infographics — not generating Ghibli-style anime portraits of oneself, which is just an one-and-done novelty.
March 26, 2025 at 4:00 PM
Pretty crazy that we've finally arrived to this point in time. Image generation is heading towards the usable and practical path with OpenAI's latest GPT-4o image gen tool.
#chatgpt #openai #imagegen #gpt4o
TIL: New OpenAI GPT-4o Image Generation
This new update by OpenAI is crazy. Finally image generation is heading towards the usable and practical path.
www.pschua.com
March 26, 2025 at 3:38 PM