LLMs are going to infiltrate all layers of software from compilers to databases to apps.

People think software engineering is dead, but honestly we’re just getting started

starting a greenfield project at work is ridiculously easy thanks to AI tooling

writes all the unit tests for you and you can just tab complete a million times..

My first question to OpenAI’s new deep research that is a PhD level researcher powered by the new o3 model is:

“How do I fine tune DeepSeek r1”

I’ve heard people sometimes refer to OpenAI’s meteoric AI progress as a consequence of possessing the “mandate of heaven”.

DeepSeek’s CEO Liang Wenfeng intentionally took aim at that very notion by open sourcing R1.

DeepSeek is quickly falling off in web search traffic while ChatGPT remains steady.

Wondering if the rate limiting turned off users or whether the novelty quickly wore off.

DeepSeek’s poetry is something else.

Can’t tell if it’s ChatGpt fatigue or if DeepSeek’s evaluation by experts outside of computer science made a big difference.

A lot has been said about App Store rankings but web search growth for DeepSeek is also stunning.

the poets and the parrots

a tale as old as time

Not financial advice, but did DeepSeek just pop the supposed AI bubble?

looks like the next global frontline for a major geopolitical conflict is our smartphones..

the reported overnight implosion of silicon valley because of deekseek r1 is overblown in my view.

this is actually a massive opportunity for the average startup. only a few are asking the right follow up questions:

DeepSeek app hit number 9 on the App Store.

A stunning rise in a market dominated by ChatGPT for so long. I just hope people are practicing good opsec when using the app.

thinking about how openai arranged a 500 billion dollar investment for future models while a small lab spent 5 million as a side project to match the performance of their flagship model..

Hot take: we need more programmers not fewer in the future. There are so many novel things left to build.

Tech layoffs are a short term phenomenon.

insane that deepseek R1 is as good as o1 and is open source with an MIT license.

The subsequent iterations of o1 models are supposedly so good that the government needed to be briefed, and deepseek out here just open sourcing R1 like it’s nothing..

Behind the Curtain: Ph.D.-level AI breakthrough expected very soon Super-agents could make AI a true replacement for human workers.

Breaking news from Axios indicates that OpenAI is gearing up for a big launch — a PhD level agent.

They give a few examples, but it seems right now it might have access to web search and a coding environment. But my guess is it’s otherwise just o3 maybe trained with improved long term coherence.

waiting for the insane chaos breaking out online if TikTok goes dark tomorrow.

Visiting New Jersey and shocked that it’s illegal to fill your own gas here ???

Found out today that the world’s tallest tree, Hyperion is 6 hours away from Palo Alto, where I live.

But it’s illegal to look for it?? What

Nvidia gets almost half its revenue from just 4 customers. Here's who they might be Some of the world's top tech companies have spent billions of dollars on Nvidia's AI chips. Its biggest customers are anonymous in filings

Kinda funny that after almost two decades of value accruing to a handful of tech companies that build applications on top of semiconductors, they’ve all decided to return that value to the semiconductor industry.

Disagree with this take. If the scaling laws hold, and intelligence levels of bigger models are unbounded, there will always be demand for higher tiers of intelligence.