14 Comments
Feb 25, 2023Liked by David Attermann

I really enjoyed reading this article, I am a farmer and user on the Threefold Grid and I could not have offered a more appropriate characterization to our focus being on the technology and growing natural demand for the token based on demand for our product.,

As community member I have created a good number of tutorials and guides and host a podcast offering help to the community with learning to use the grid.

If you have any interest in taking a deeper look into what we are doing, I would be happy to orchestrate some time for you to be able to have free reign of Q&A, with either myself or a staff member by your preference.

It’s often hard to see the full scope of our project but I did want to share a couple links that help reign it all in

Threefold is absolutely as open source and inclusive as they come buts it a very large project and even just the core team itself includes over 70 developers.

Our project’s primary repo is here

https://github.com/threefoldtech

The repos for our website and manuals are here

https://github.com/threefoldfoundation

There is a full “official” manual located at

https://Manual.grid.tf

My community manual lives at

https://deploy.threefold.dev

My channel where I host recordings of the Q&A style calls aswell as video tutorials is here

https://youtube.com/@drewsmith5675

Hopefully we can all build a better digital future together!

Expand full comment
Feb 24, 2023Liked by David Attermann

If you want to add utilization of TFT to the chart, check the newest posts here

https://forum.threefold.io/t/grid-stats-new-nodes-utilization-overview/3291/42

Expand full comment
author

this is great, ty!

Expand full comment
Feb 24, 2023Liked by David Attermann

The most well researched topic on this subject ever put together. Note TFT is on XLM and BSC, not DOT.

Expand full comment
author

appreciate the kind words! isn't TF chain a parachain though?

Expand full comment
Feb 24, 2023Liked by David Attermann

TfChain , which currently runs on Parity Substrate, was going to move to parachain. They are having second thoughts and may chose something else, possibly Algo. Currently, tokens are moved to TfChain when being used to deploy a VM, but farmers earn on XLM. Once we move blockchains and launch validators, earning and deployments will be on the new chain. There is currently no trading on TFchain, only on XLM and BSC. I'm assuming those will remain tradeable in the future.

Expand full comment
author

ah great color, ty ser

Expand full comment
Apr 10, 2023Liked by David Attermann

What are your thoughts on Phala network (PHA) and where would they fit in to the spectrum of decentralized cloud services?

Expand full comment
author

at first glance seems to fit within the fungible compute group since workloads are executed across clusters. will have to do more work to have a real opinion on it tho. ty for bringing to my attention!

Expand full comment
Mar 6, 2023Liked by David Attermann

Hey man, great piece and I love the fresh take and framing for this category. Curious to know whether with your analysis you took into consideration existing off-chain compute networks that have already demonstrated some degree of PM fit. Projects like Gelato come to mind, or even indexing/querying protocols like the Graph which are arguably performing specialised and cryptographically verifiable computations off-chain.

Also I'm struggling to really gauge whether the incentives for developers to monetise via microservices will outweigh the financial upside from forking existing dApps to death. Possible to provide more colour on how you see this unfolding?

Expand full comment
author

much appreciated! never heard of gelato but the project looks interesting -- sounds like chainlink's automation product. not sure if its suited for more complex compute but will def do more work on it. in terms of projects like the graph -- I view them more as specific services vs. general purpose compute.

obviously this is all speculation, but I think the incentive system can change with this new microservices capability. logic being talented devs will want to "IP protect" their code (others can still use but they cant monetize) and this should create apps with superior UI/UX and utility. sure anyone can fork old code and make a copy cat app -- but in the long-run I think these will all die out since they have no innovation / incremental utility.

Expand full comment

What's your critique on the Coinbase's most recent AI Research: Crypto's AI Mirage? It outlines the short to medium term challenge by centralized competitors that for DeComp. The centralized counterparts have amassed cutting edge tech over the past 2 decades.

Expand full comment

i'd point you to a piece i did a year ago on the value prop of web3 ai (https://omnida.substack.com/p/blockchain-and-ai-sum-much-larger)

in short, i believe the permissionless & composable nature of web3 ai will enable novel technical functionality & business models not possible w/ web2 networking architecture.

think the cb piece kinda missed the mark; the point isnt to be at "parity" w/ centralized ai models...the real value imo will be connecting smart contracts w/ LLMs (most people dont care about decentralization - but they will pay for useful and cool utility)

while i do agree w/ cb that there is much work to be done (across all web3), i think TEE & MPC (like https://aizelnetwork.com/) will connect trustless off-chain LLMs w/ smart contracts much faster than most expect, and much more practically than zkML (which is likely 5 years from being feasible for most use cases).

Expand full comment

I think I'm gradually starting understand your composability breeds utility thesis more and more as I read you article back and forth more. To help me understand a bit more, would you be able to expand more on the topic you mentioned below? What are some real world projects currently be worked on to achieve this?

"Composability is no longer confined to blockchain compute: Trust-minimized compute projects are essentially oracles that connect blockchains to off-chain compute. This opens up a massive amount of compute power, since each node can incrementally contribute to the network’s capacity (vs blockchain nodes that must all process the same state). The result is composability with (theoretically) unlimited compute capacity."

Expand full comment