Mike (Light)

- @lightterminal -Two exits in non-crypto

[Mike (Light) @IdeaMikk]

⬇️⬇️⬇️

[Light @lightterminal]

A new way to trade together.

Catch runners early, automate trades, get paid for alpha.

[Mike (Light) @IdeaMikk]

@ivanfromtheweb @solanamobile 🤝

[Mike (Light) @IdeaMikk]

@Autismo_Beanz @lightterminal hey 🤝

[Light @lightterminal]

Fortune 500 companies are endorsing Solana projects and you're bearish?

$THREE

[Mike (Light) @IdeaMikk]

@Dreamonology @AutismCapital It's the same in Europe. People say "you vote for them" but in reality very few do, but somehow the same people who've ruined the continent end up getting the most votes.

[Mike (Light) @IdeaMikk]

@zeroxkyle Yeah, I think about it a lot. There are many people who can pick the right assets and ride the port up, but few who know how to get out of the trades while being up.

[Mike (Light) @IdeaMikk]

@itzmac_ @lightterminal I like you even more now

[Mike (Light) @IdeaMikk]

@0xSadGhost Not my game atm :)

[Mike (Light) @IdeaMikk]

@itzmac_ @lightterminal Yeah. Non-slop takes time, also when bootstrapping

[Mac @itzmac_]

extremely good tek btw

@lightterminal

[Mike (Light) @IdeaMikk]

common mistakes by builders that I learn every day:

- building a great product isn't enough, distribution is everything

- a great product alone doesn't make anyone to change what they're already doing if they're more or less satisfied as is

- to have a chance at building something really big, you need both non-consensus insight about the market and inflection point (currently it's AI). Good book to read is "Pattern Breakers".

What is your non-consensus, unique take on the market that only few people understand?

- you can still build a good business by better execution, better distribution, or having more money to stay alive longer/spend on acquisition, but it won't be a product that shapes the industry.

- good team alone isn't enough. You need a great team who's also hungry and resilient and aligned. Also, people have different seasons in life. Someone very capable in a less productive season of life (becoming a father or going through some other personal thing) isn't the same person who he or she is in a more productive season of life.

- a good team doesn't mean the project valuation will go to many millions or the product will succeed. This is maybe more for investors.

- targeting the right audience early is everything. The early feedback will shape the product the most.

- there are many people who have a good talk but little substance. Meaning, they're used to "vibing" as this has worked for them in the easy times. In hard times, it's not only useless, but it's harmful hopium of results that aren't coming bc the person won't put in the work.

[Mike (Light) @IdeaMikk]

@_proxystudio @MCGlive @m00npapi yeah mcg is dope

[Mike (Light) @IdeaMikk]

@nerdy_hex Or zig when others zag?

[Mike (Light) @IdeaMikk]

AI coins doing well even when BTC dips. Tells you all you need to know.

[Mike (Light) @IdeaMikk]

@ThePumponomics wait, this is actually interesting

[Mike (Light) @IdeaMikk]

@razeaii good work man, congrats on the launch

[Mike (Light) @IdeaMikk]

@KRinweb3 @lightterminal 🤝🤝

[Tommy @Shaughnessy119]

I'm cautiously optimistic we are seeing the early signs of a new bull market in crypto, but very different from the last few

Foundationally we have the Clarity act and CFTC/SEC tripping over themselves to allow Crypto but we’ve historically been missing an animal spirits demand driver outside of Saylor for BTC

Hyperliquid doing extremely well is the main center of gravity especially as they launch non crypto markets (Oil, Gas, Pre IPO stocks).

We’re also starting to see Crypto x AI tokens doing well which has always been the largest potential new sector imo. I’ve talked about this a lot in recent tweets.

Venice has driven a ton of excitement on the AI inference side and this should continue for four reasons 1/ AI inference demand is infinite, 2/ business want to cut AI costs and can pay 1/100 the cost using open source models 3/ eventually agents will need to hold tokens like DIEM to access inference on their own and eventually grow their holdings to access more intelligence and 4/ NSFW and use cases the labs dont allow. You can argue whats legal/moral but its not up to Dario to tell you what you can ask

I think the distributed AI Inference trend ends if Chinese models go closed source but I don't see that anytime soon. Also the inference providers (OpenRouter, Venice, etc) should buy GPUs to lower costs further and use them to train/fine tune models in the future if the world goes closed source. How else does OpenRouter use $113 million U.S. Dollars?

Outside inference we’re seeing a wave of new AI projects/tokens and adjacent ones do well (Nock with merged mining to reuse AI inference to secure its chain, Pearl hitting $2b+ out of the gate (low liq, OTC), and a huge swath of new AI projects on base that are more infra and legit vs the last swath of reply bots with tokens). I'm seeing numerous inference providers on Base/Solana compete to the cost of electricity for inference which is useful vs the last era of AI reply bots competing for likes, this is better. Grass is also doing extremely well revenue wise selling to AI companies.

I think people will really lean into using their @NousResearch Hermes agents to build the agentic economy in crypto (in addition to the huge traditional sectors they are targeting). This starts with giving your agent a wallet for basic transactions, grows into them interacting autonomously and the final step is agents creating their own AI economy of protocols and DeFi services. You’ll click a button to graduate your agent to autonomous mode eventually and it’ll be wild. I still feel folks will want this run on a box inside their home especially if it has 24/7 access to your life (cams, mics, all data).

The next potential bull market in Crypto will be different for a few reasons though

1/ Every investor has the choice of buying AI stocks, private companies and deals vs Crypto/tokens so the opportunity cost has gone way up. I'm not investing in a crypto project if I see a better AI one for my money. This extends to retail who has to decide if they want to buy uncensored money ($BTC) or AGI (OpenAI/Anthropic stock) with their marginal dollar. TLDR mid crypto projects never get a shot now, nor should they.

At Delphi Ventures we are backing early stage AI and early stage crypto founders

2/ There is a bull market in Wall Street for Blockchain. Every company has a stablecoin, or integration or an ETF. This is fantastic and what we all wanted but is separate from the on-chain animal spirits world we all look at. When I read our year ahead report for Infra I was shocked and excited with just how embedded crypto has become within Wall Street, and how important stablecoins are for our the united states (sell the dollar via Crypto rails globally so you can lower borrow rates and embed it into new economies)

3/ The people want real products, they want revenue and they want that revenue to flow to the token. No one cares about altruistic academic circle jerk conversations. It's not 2018.

Because of the above, this cycle will become ultra concentrated in a few winners

Why? Because there are only a few projects that have PMF as a product, drive revenue, and rise above the opportunity cost of folks to buy other assets (in Crypto and in AI).

If you are an early stage founder who is way out of distribution building in AI or in Crypto and are raising at the earliest stage reach out. A warm intro is the best path.