Peter Kris

Peter Kris

2,489 Twitter followers

Follow

cofounder @gasp_xyz / defi / mev / L2s / AMM maxi / onchain connoisseur

Posts

let me coin the term - appfilm and AI agent workflow designed to harness human workforce through a custom movie generation engine. It will start as avatars over voice agents (e.g. doing HR interviews), but once it transitions to more managerial role, you can add more spice to movie-fi the whole work process. Fun future

yeah, the transformation happening now is deep, this is a roughly once-in-40y change, so pretty deep. If you don’t share this view, or don’t understand what I’m talking about, just be prepared for surprises

first check $500k-1M pre-seed

i’ve never seen such well funded startups become irrelevant so quickly. we are living in unprecedented but dangerous times for VCs

I always thought that gaming and crypto works well together because both are substantially *virtual* It turned out, if you financialize a game, it looses it's appeal. No one wants real world consequences in a virtual fun game

So can we all admit that play to earn was the dumbest thing of all time and gaming on da blockchain was a tremendously stupid idea through and through x.com/degeneratenews…

absolutely The LLM is a somewhat lossy compression of snapshot of humanity’s formal semantics. Every prompt is a decoding key for decompression Beyond the buzzwords - most of humanity culture can be expressed by a small subset of weights. Thats saying that human culture is rather simple and unoriginal

yes and no the dead internetization is killing internets, and there is a trend of offline AI is homogenizing our global culture, yet people are forced to specialize outside to have any *edge* we're heading to a weirder future with more global internet with more power-law like products/trends combined with new local offline siloes, which are creating information frictions by definition and by desire

so now we can have LLM inside an LLM ? no thats not MoE

Christos Tzamos

@ChristosTzamos

1/4 LLMs solve research grade math problems but struggle with basic calculations. We bridge this gap by turning them to computers. We built a computer INSIDE a transformer that can run programs for millions of steps in seconds solving even the hardest Sudokus with 100% accuracy

Everyone from AI executives and researchers to doomsayers, politicians, and activists is talking about Artificial General Intelligence (AGI). Yet, they often don't seem to agree on its exact definition. One common definition of AGI is an AI that can do everything a human can do, but are humans truly general? In this paper, we address what's wrong with our conception of AGI, and why, even in its most coherent formulation, it is a flawed concept to describe the future of AI. We explore whether the most widely accepted definitions are plausible, useful, and truly general. We argue that AI must embrace specialization, rather than strive for generality, and in its specialization strive for superhuman performance, and introduce Superhuman Adaptable Intelligence (SAI). SAI is defined as intelligence that can learn to exceed humans at anything important that we can do, and that can fill in the skill gaps where humans are incapable. We then lay out how SAI can help hone a discussion around AI that was blurred by an overloaded definition of AGI, and extrapolate the implications of using it as a guide for the future.

AI Must Embrace Specialization via Superhuman Adaptable Intelligence

Everyone from AI executives and researchers to doomsayers, politicians, and activists is talking about Artificial General Intelligence (AGI). Yet, they often don't seem to agree on its exact...

We have something brutal ahead of us 1. Block has annouced layoffs affecting 40% of workforce 2. Stock price surged 24% in minutes Other companies will take this as a blueprint to save their falling stock prices. Be ready for storm and high unemployment

I'm clamoring the end of scalability L2s for some time and now it's coming to reality A bit sad - so much capital and R&D was spent on this. Customized L2s that allow new tech stack, or app-specific logic or expanding features are the only ones that makes sense. (looking forward to OP_CAT and Bitcoin L2s later on in adoption cycle)

@VitalikButerin

There have recently been some discussions on the ongoing role of L2s in the Ethereum ecosystem, especially in the face of two facts: * L2s' progress to stage 2 (and, secondarily, on interop) has been far slower and more difficult than originally expected * L1 itself is scaling,

A friend wanted to invest in HYPE, so I made a quick analysis. Sharing here, if it's useful to someone (NFA, DYOR) : pros - 1.⁠ ⁠it’s rare and usually a great sign when in bear and bad times something grows (in price or rev). This was great indicator over past many years. E.g. Ethereum ICO has raised the 2nd biggest internet crowdfunding in 2014 bear market. 2. they do smart regulatory arbitrage cons- 1.⁠ ⁠every 2y you have a perp exchange. Started with DyDX, was replaced by GMX, was replaced by Hyperliquid. Why would someone not replace Hyperliquid again? 2.⁠ ⁠their customer base is less-sticky. Gamblers loose money and go away, whales centralize wealth and move elsewhere. Trader customer base is harder to retain. 3.⁠ ⁠they are just now running into scaling bottlenecks, bc they want to truly sell themselves as Binance contender, but it will be impossible due to their decentralized nature 4.⁠ ⁠numbers can lie, the price of HYPE might be manipulated as it originated and most volume is done on Hyperliquid itself. 5.⁠ ⁠⁠Numbers are lying about their liquidity depth, they’ve implemented preferential cancels, where you can super easily spoof limit orders, and then just post cancel when you see something happening against you. It’s easier to fake/spoof limits Conclusion: the overal makes me actually neutral, with sometimes leaning slightly negative, which makes me too un-excited or lazy to move finger and funds to invest, therefore not investing

Loading..