Comments URL: https://news.ycombinator.com/item?id=47943978

Points: 1

# Comments: 0

If Peter Steinberger is able to generate even a 100M this year from Clawdbot what he has is a multi billion dollar business that would be life-changing even for a successful entrepreneur like him who is already a multi-millionaire. If it collapses from the security flaws, and other potential safety issues he loses nothing, starting from zero and going back to it. Peter Steinberger (and startups in general) have a lot to gain and very little or close to nothing to lose.

The iPhone generated 400B in revenue for Apple in 2025. Clawdbot even if it contributes 4B in revenue this very year would not move the needle much for Apple. On the contrary, if Apple rushes and botches releasing something like this they might just collapse this 400B/annum income stream. Apple and other large enterprises (and their execs) have a lot to lose and very little to gain from rushing into something like this.

Anthropomorphizing removed, it simply means that we do not yet understand the internal logic of LLM. Much less disturbing than you suggest.

2) Most LLM or agentic applications don't give us enough control over the process. Investment research is pretty open-ended and subjective, there is no one absolute approach. Individuals and teams often have some tribal perspective on how to do it or needs certain ground covered to have conviction. Existing solutions do not allow such control.

I am a first time founder working on Vector -- an agentic platform for investment research and analysis. Looking for feedback on the public beta that was launched, specifically around the unique philosophy behind this product. Which usually ends up either really gamechanging or just dumb -- I recognize this and hence feel the need to try and validate things as early as possible.

This product was born from personal needs and frustration with existing platforms and tools. I am a retail investor myself managing a significant personal portfolio. Even though I only rebalance quarterly, it takes significant time and effort to keep up with emerging risks to my portfolio and even more time to research and develop conviction around new ideas for diversification. Seeking Alpha is a chore to go through with too much generic content, inconsistent quality and noise. ChatGPT and Perplexity Finance lack a decent workflow (too chatbot centric). Deep research might seem great at first, but the lack of control and verbosity becomes a pain very quickly.

Vector is designed around two core theses that we believe differentiate it meaningfully from existing solutions

1) User-in-the-loop. We recognize that AI agents still limited in reasoning capabilities and require guidance for many complex tasks. Secondly, we also believe that investment research is an open ended and rather subjective task. Useful research processes(ones that help consumers achieve/overturn convictions)require the means to let the consumer change/challenge the methodology and also the assumptions made.

2) Embed agents into well tested workflow, not force a chatbot on users. Many AI/LLM powered products out there today are chatbot first for some reason. Some claim that this is because natural language is the most powerful HCI given the capabilities of LLMs today. It is true in some instances but false for many others. For the problem we are tackling, a chatbot is not a natural workflow for investors and creates a lot of friction. Classic concepts like watchlist, stock screeners and trending news are actually well tested and very sensible. They do have limits, but that is where embedding agentic capabilities adds value instead of serving as gimmicks.

We’re releasing this early, imperfect beta not because it’s finished, but because the core theses behind Vector deserve to be tested. We’d love your feedback — even (especially) the critical kind — to help shape what comes next.

Public beta (Free): https://vector.deepinsightlabs.ai Discord server: https://discord.gg/CKzKWaex Website: https://www.deepinsightlabs.ai

P.S I would remove the sign-up and login for easier access, but the watchlist functionalities require unique identity for users to work. Apologies for that.

Comments URL: https://news.ycombinator.com/item?id=44695841

Points: 3

# Comments: 2

I am dropping by NYC as a side trip about two weeks from now and am looking for tips and advice on forming connections with the startup community, potential clients and investors there.

Specifically interested in the wealth management space (independent financial advisors, external asset managers), hedge funds and asset managers as well as other startups and VCs that are in the same space.

Doing this with the intention to enter the U.S market within the next 6 months, i.e GTM hire and starting sales motion and pilots. Any suggestions on where and how to get started?

https://www.deepinsightlabs.ai

Comments URL: https://news.ycombinator.com/item?id=44695402

Points: 2

# Comments: 4

This is why I always believed that VCs and accelerators that mislead startups about how early they are willing to invest are doing a lot of damage to gullible founders.

This is why knowledge management is such a popular use for POCs involving LLMs, and ironically also why POCs don't progress into something more permanent

Has anyone tried this out? Curious what the technical implementation might be and what future use cases this unlocks (e.g. automated 10-K comparators, anomaly spotting, regulatory change detection).

Would love to hear thoughts from professionals and serious investors, as well as technologists working in the space.

Comments URL: https://news.ycombinator.com/item?id=44193675

Points: 1

# Comments: 1

Comments URL: https://news.ycombinator.com/item?id=43752633

Points: 1

# Comments: 0

Asking the LLM to think step by step simply biases it towards the latter. It's still a stochastic parrot but now it sounds logical and that happens to be useful in some cases, regardless of whether we can agree if it's "reasoning".

Comments URL: https://news.ycombinator.com/item?id=43275143

Points: 1

# Comments: 0

At least until compute cost drop to a cheap enough level...