Alex Atallah

Made a way to pay for LLM credits with crypto. Interesting how few platforms there are that do crypto payments well: https://twitter.com/xanderatallah/status/1686754529037996032?s=46

Has anyone here used loopcrypto.xyz for crypto payments before?

Which companies have regular hackathons scheduled multiple times per year? How sustainable is it?

Decided to open-source Window AI. Some ideas if you’re looking for a fun weekend hack (no api key required): - control MetaMask using AI - a game generator, make your own games w natural language - ideas and templates on skylightai.io Docs and code: https://github.com/alexanderatallah/window.ai

Launching an experiment this week to enable pseudo-decentralized gen AI apps - lmk if you're interested in trying it out and providing early feedback

What if some non-human species evolved to the point where its members could speak English, use the internet, and express a demand to be treated equally with humans.

Is anyone doing to Hoon what Elixir did to Erlang?

Loving the Unix principles underneath Redis - inspired thought: https://twitter.com/xanderatallah/status/1617576326432722944

Has anyone else here gotten weirdly into Redis?

Pubsub as a philosophy for consumer tech: https://i.imgur.com/CASjicG.png

Super interesting visualization of a neural network, particularly for seeing the perspective of individual neurons: https://playground.tensorflow.org/

“To consciously activate your DMN and creative ideas during the day, allow yourself to spend time doing activities that aren’t cognitively demanding… without listening to music or a podcast. Simply let your mind wander.” https://www.bespacific.com/science-of-why-you-have-great-ideas-in-the-shower/

People with a personal blog: what’s the link, what tech/product stack did you use, and why?

Will AI training bifurcate into pre-training and prompt engineering? or will there continue to be one level (or more) of fine-tuning in between?

Both crypto and AI seem to involve thinking "in the limit." Crypto: what will people do with these incentives in the long run? And how do we solve cold-start in the short run? AI: how will bots perform with these parameters in the long run? and how do we tune them as gradient -> 0?

Post-ChatGPT, what will happen to open source business models that charge for consultation and example code? TailwindCSS example: https://i.imgur.com/7GZRnj2.png

NFTs are the easiest way to make your own business. $1b paid in creator fees this year: https://twitter.com/opensea/status/1597984856369074176?s=46&t=_yfYUwFy476a30ykdh2mIg

Assumption: there are no truly original ideas. Everything is a remix of something old and something new. Great ideas just stem from surprising combinations of both. So here’s the big q: will LLMs like GPT-3 become better at doing this that we are?

Better to have great execution but going the wrong direction, or have worse execution but in the right direction?

It's clarifying to identify and question the most important reason for doing something, instead of searching for multiple. If there is no single reason for doing it, it may not be important at all.