Algorithm tweak coming soon to promote more informational/entertaining content. We will publish the changes to @engineering.
Our goal is to maximize unregretted user-seconds. Too much negativity is being pushed that technically grows user time, but not unregretted user time.
First, black people were discriminated against, and became resentful.
Then, DEI and affirmative action were invented.
Thus, white people were discriminated against.
Now, white people are resentful too.
Discrimination does not work.
It just transfers resentment and creates more racists.
If you go far enough right or left you’ll find resentful racists.
They’re just different colors.
my 2025 crypto developer thesis
as a developer in 2025, you have 3 options
- SVM
- EVM
- Move
EVM will be by far the most competitive: Hyperliquid, Monad, Berachain, Base, Megaeth, and many new EVM L2s
SVM will be predominantly Solana and a few other L2s (Eclipse and Atlas most notable) + perhaps one SVM L1 fork akin to Pythnet
Move will be Sui, Aptos, and Move L2s
few consequences of this:
i) EVM teams will be more mobile. If their existing EVM ecosystem is not meeting their needs, they'll be able to move to a newer EVM ecosystem quickly and with little friction
meaning I expect all EVM ecosystems to compete against each other hard this year — and since the dev experiences will be relatively similar, the moat will have to come from other avenues (ecosystem support, liquidity, and "community")
you could also cross-deploy on multiple EVM environments of course and some will, but this becomes complex to manage and your product generally suffers in focus
ii) SVM and Move will have an edge for developer stickiness and talent density because those devs won't have many options to choose from, they will have a stronger incentive to stick it out
this is also why back in the day I spammed "only possible on solana" (OPOS) — the reason is that the platform can give you a head-start edge since copy/pasting EVM forks won't work in either direction
iii) however, since there are many more EVM ecosystems, I would expect the EVM dev tooling to advance more rapidly, generate more data for LLMs, and overall offer a better developer experience due to the abundance of educational resources
iv) at the same time, since the EVM is more mature in years, I suspect there are fewer low-hanging fruit for progress and progress might be slower, relatively
v) which brings me to my main point: the SVM and Move ecosystems have an inherent technical edge in 2025 due to lessened competition, higher talent density, and stronger incentives for devs
BUT, this will be entirely dependent on one key factor: how fast the SVM and Move ecosystems advance their respective developer experiences — contract-level, read-level, and core protocol-level
meaning, for example, if Solana can improve its developer experience 2-5x in 2025, its growth for the year could very well be 10x relative to others
but if it doesn't, this will be a huge setback — there can not be another "congestion" episode
as a result, @heliuslabs will focus 100% of our efforts on improving the SVM developer experience without a single day off in 2025 — if you're curious about building on Solana or other SVM L2s, give me a shout
let the dev tools arms race begin