Vincent Chen

Announcing @LeadbayAI $4.3M to kill everything you're doing with sales. Corgi, Deel, L'Oréal recently told us: "We found in 10 min on Leadbay what we couldn't in 6 months with Clay and a GTM agency." Leadbay qualifies the millions of data-signal-scarce companies nobody could find before. This is how it works:

@mirandanover I want this !!!

Introducing Fort, a wearable that automatically tracks strength training. Strength training is one of the best things you can do for your health and longevity. It deserves better tools.

@ycombinator @cofia_ai @paomartcard @_moseswayne Woooooh!

.@cofia_ai creates AI automations that write themselves. They learn how you work, and proactively deploy tailor-made automations without you ever writing a prompt, coding, or building a workflow. Congrats on the launch, @paomartcard and @_moseswayne! ycombinator.com/launches/Pbk-c…

hatetocall .com just one-shotted rescheduling a dentist appointment by phone for me. Unreal.

@gauravvohra Checkout hatetocall.com I used it to negotiate with my car insurance, saving me 700 dollars in one call

ISO AI agent to make ad hoc phone calls on my behalf Example use case: reschedule doc apptmt Recommended product that does this?

@blakeandersonw That is inspiring my Blake

My “successful” businesses - ai apps My “failed” businesses - ebooks - RuneScape botting - madden mobile coins - Minecraft servers - eth mining machine (2017) - drop shipping - instagram meme pages - social media marketing agency - youtube - online tutoring - amazon fba - landscaping - local it assistance - self improvement brands (apex + optisci) - reselling - ai apps

After shutting down my YC company, I spent >6 months looking for a job and grinding leetcode alone, despite having interned at @nvidia and @amazon. So I built an AI to do live mock coding interviews with and finally landed a great role. Hopefully I’m the last solo leetcoder.

@nrehiew_ Hey do you manage to find any?

Does anyone have working code for GRPO + LoRA at some decently long sequence length. I have 2xA100 OOMing with 4096 + sequence length, 8 rollouts, rank 8, DeepSpeed 0.15 even without vllm It needs to be at least >4096 because the distill models have CoT of that length for problems of any meaningful difficulty

Very soon long context agent’s inference cost will drop several times with these sparse attention

Today we’re launching the Anthropic Economic Index, a new initiative aimed at understanding AI's impact on the economy over time. The Index’s first paper analyzes millions of anonymized Claude conversations to reveal how AI is being used today in tasks across the economy.

3 months to turn your girlfriend into a stay at home pilates mom in a Porsche Macan

browser-use ftw

@andrenuyens_ Interested!

For anyone in SF, I am putting an extreme poly-athletes group to do crazy adrenaline sh*t together. We’ve got skydivers, hikers, bikers, surfers, kitesurfers, paragliders, divers and more between all of us We are 12 rn. If that’s you, DM me!

@daltonc Sedgwick

If you think of Instacart as a do-over of Webvan, what other companies are ripe for a do-over?

@ed_delan @andrewchen Might be similar to what you describe?

@ed_delan @andrewchen I join the discord as a user, and I love the thing! Supposedly they are beta testing the ability for for you to interract with the ai podcast host and guest in realtime and record it

Does anyone have a reco for a podcast co-host/interviewer AI app? - you give it an essay / deck / topic / whatever - it asks you questions, and follow-ups based on what you say. It knows when you’ve said what you want to say (your answer gets short or your answer is less interesting or repetitive) - sometimes makes observations or expands on a topic, to transition topics - it can grab quotes, excerpts, and ask you to react - bonus if it does nice things with the output (editing for pacing etc)

Adding intermediate reasoning tokens effectively adds layers to the computer graph, making the model deeper and allowing it to simulate deeper circuits. We might soon see LLM can potentially be Turing complete

Can neural networks be Turing complete? Traditionally, without CoT, transformers are limited to parallelizable tasks within the AC0 complexity class, solving problems computable by shallow Boolean circuits due to their fixed architecture and inherent parallelism. CoT reasoning fundamentally changes this landscape by enabling transformers to handle sequential computations through intermediate reasoning tokens. Each CoT step effectively adds a layer to the computational graph, increasing computational depth and allowing the model to simulate deeper circuits beyond AC0. This advancement brings transformers into the realm of P/poly—the class of problems solvable by polynomial-size circuits. With sufficient CoT steps, a transformer could, in theory, simulate any computation that a polynomial-size circuit can perform, narrowing the gap between transformers and Turing machines. Practical limitations exist, such as finite context windows and computational resources. Leveraging this potential requires careful model design and optimization. Extending this concept, we can envision neural networks functioning analogously to operating systems. Just as an OS manages resources and orchestrates complex processes, neural networks could dynamically control computational tasks, allocate attention, and sequence operations to solve intricate problems efficiently. Integrating CoT reasoning in transformers offers a promising path toward Turing completeness. Despite practical challenges, this perspective could revolutionize artificial intelligence by transforming neural networks into general-purpose computational frameworks capable of complex reasoning and decision-making. An excellent paper

@rosaticorp I need some pilot founder friends

Entrepreneurs who are pilots = client lunch meetings in a different state, home for dinner 🙌🏻 stoked to tag along

Add me to your group chat if you are in SF