Deano 💹🧲 (@DeansCleanBean) header
Deano 💹🧲 (@DeansCleanBean) profile picture

I degen, you degen, they degen, we degen. I only degen the best degens

Joined November 2019
Photo by Deano 💹🧲 (@DeansCleanBean)Photo by Deano 💹🧲 (@DeansCleanBean)Photo by Deano 💹🧲 (@DeansCleanBean)Photo by Deano 💹🧲 (@DeansCleanBean)Photo by Deano 💹🧲 (@DeansCleanBean)Photo by Deano 💹🧲 (@DeansCleanBean)Photo by Deano 💹🧲 (@DeansCleanBean)Photo by Deano 💹🧲 (@DeansCleanBean)Photo by Deano 💹🧲 (@DeansCleanBean)Photo by Deano 💹🧲 (@DeansCleanBean)
Deano 💹🧲 retweeted
"When I was small, not much bigger than a pollywog," said Frog, "My father said to me, Son, this is a cold, gray day. but spring is just around the corner."
8
131
13
1,427
18,728
Deano 💹🧲 retweeted
We can announce today that we have been chosen by “The Hitchhikers Guide to the Galaxy” as the core provider of AI agent technology in a multi-year, multi-event campaign. Selling over 14 million copies, "The Hitchhikers Guide to the Galaxy" is one of the largest and most influential franchises in the world, responsible for shaping today’s internet culture. This collaboration brings "The Hitchhikers Guide to the Galaxy" IP into the age of AI. We will be integrating SERV’s AI technology into the Hitchhikers' universe in an extensive series of real life events, software, multimedia integrations & token launches, all as part of a long term, multi-year project. This collaboration is going to bring a lot of exposure to our technology in the mainstream, with high-profile personalities involved, PR and exciting globally renowned partners on board, all of which is going to significantly increase awareness of what our technology offers to companies - and take the whole platform up to the next level. More information will be shared as we begin development on this important project. 42.
81
98
37
399
127,935
Deano 💹🧲 retweeted
Its hobaning
7
10
1
25
543
Deano 💹🧲 retweeted
No buyers No sellers No cares $hoba forever or whatever, idc
5
7
17
273
Deano 💹🧲 retweeted
Thesis remains the same: we dont care
7
8
31
543
Deano 💹🧲 retweeted
I love him too much man
7
413
15
2,868
47,844
Deano 💹🧲 retweeted
Elite chant man wow
7
417
22
5,002
91,115
Deano 💹🧲 retweeted
LIFT IT UP ©️
2,036
54,002
2,583
162,686
3,028,296
Remember it like yesterday! It’s going to be sooooooo good to finally see this trophy dressed in Arsenal colours again! I know the lads and staff will enjoy every single, well deserved minute of today’s game and there’ll be a party on the streets of london tonight!! #coyg #londonisred #arsenal #northlondonforever ♥️🤍🏆
68
1,629
9
12,869
114,720
"When we do win, it will be down to everyone else to deal with that" Ebere Eze does not care about the outside noise 🥶
144
1,687
133
13,970
647,417
Deano 💹🧲 retweeted
We did it, together.
6,921
93,855
7,325
388,728
14,610,028
Deano 💹🧲 retweeted
The Arsenal. Your Premier League champions.
27,598
201,589
35,303
579,597
39,688,678
Deano 💹🧲 retweeted
These are real numbers in production with an enterprise client during our first week of the SERV Reasoning Private Beta. For those who are new, SERV Reasoning is a proprietary AI reasoning API that solves the biggest issue in AI Agents today: > their ability to produce reliable results, consistently at an economically feasible price point. It is the largest bottleneck in AI adoption across startups, enterprise, and governments. The product is currently in private beta, and has already been deployed across 10+ enterprises, the UAE Government, and multiple startups in prediction markets, compliance, DeFi, private AI, banking and more. > How does it work? Instead of asking a model to think step by step in open prose, SERV Reasoning binds the model to execute a custom reasoning graph expressed in machine-native code (Mermaid flowchart syntax, for the technical readers). The core graph mechanic splits the problem into two phases: 1. A high-capacity model generates the reasoning graph. Branches, validation checkpoints, decision nodes, all laid out explicitly as a plan. 2. A smaller model executes the graph at inference time, node by node. It follows the plan. It doesn't improvise. And now, with a simple one-line code change, any production AI Agent system can benefit from this reasoning engine. We're already seeing insane results in production with our enterprise users, see quoted tweet below. This is all backed by a full academic research paper we published in December, co-authored by an NVIDIA AI researcher. It is currently undergoing peer review. And this is all just the beginning. We've got a stacked roadmap for this product with: > Shadow agents > Sharding Graphs > Morpheme-aware LLMs > Our own fully custom built models 100x improvements in performance-per-$ in production with our first enterprise users is literally just the start. SERV.
Decentralized open source models are getting very interesting Openserv “SERV reasoning” just received a strong independent benchmark Headline findings (SERV-nano solo vs prior Gemini/Sonnet): - 100x cheaper per call ($0.0006 vs $0.06) - 107x performance per dollar - Higher accuracy at lower cost (83.3% vs 77.5%) Keep an eye on the smaller teams shipping tangible products We’ve already seen VVV go on a solid run with arguably more upside, dolphin filling in for a beta position on the privacy AI narrative I’ll link a list of protocols to look into below, so bookmark this:
10
26
13
148
17,829
Deano 💹🧲 retweeted
Greg Ivanov joins SERV as key advisor. ex-Head of Partnerships at Google, GP at 227 - early backers of Pendle, Bittensor, RNDR. He grew Google Play into one of the largest developer ecosystems on the planet - now working with us to scale SERV Reasoning across large enterprises.
41
68
45
344
108,839
Deano 💹🧲 retweeted
SERV Reasoning Private Beta is live, with hundreds on the waitlist. One of the first teams - ThoughtProof - builds agent verification infrastructure for banking, compliance, and onchain settlement. They ran SERV against their production stack. Result: 107× PPD vs baseline.👇
14
58
22
229
116,597
Deano 💹🧲 retweeted
Let me lay out $SERV tokenomics in short, because I keep seeing the question. The platform has one core engine: SERV Reasoning. It's frontier-grade AI reasoning at 50x lower cost than the major labs. It's already in production for governments (UAE disclosed, others under NDA), and an enterprise pipeline filling fast. 25% of platform rev → $SERV buyback and burn. Three main flows feed it: 1) Enterprise contracts and API usage. The heavy hitter. 2) SERV Build - developers shipping agents that run reasoning calls. Every call, more revenue, more burn. 3) SERV Launch - permissionless tokenization layer. Every TGE, every fee, all routes back to main token. Now get this: developers globally spend $300-500M per week on LLM compute, and the trend is only accelerating. SERV is the cheaper, auditable layer underneath that curve - every dollar that flows through it burns the token. The L3 catalyst is a bit further out - its when reasoning moves on-chain, and every API call settles in $SERV directly. The interest so far is immense (btw I just heard some incredibly bullish feedback from beta users), so the public API opening should be wild. It doesnt get any more obvious.
29,638 requests and 279M tokens in week one of SERV Reasoning private beta. API is in the hands of developers across DNA sequencing, autonomous penetration testing, satellite imagery, and edge AI for transportation. Hundreds on the waitlist - batch 2 invites going out shortly.
13
31
9
116
16,280
Deano 💹🧲 retweeted
Thanx @rekt_tekashi Newest addition to me memecoin plushie collection BOBO the bear, and good luck with ur migration and all that. Also pictured is GONDOLA APU HARRYPOTTEROBAMASONIC10INU and HOBA (the greatest honeybadger themed token ever of all time ever)
17
24
2
71
2,395
Deano 💹🧲 retweeted
Imagine a $23M FDV crypto company is being used by the Government of the UAE, a Government with $1.7T in sovereign assets and going all in AI. @openservai isn't in pilot at the Government of the UAE, it's live in production right now. The opportunity to research $SERV early is now, while most of CT is busy vibe coding and still hasn't done the research. Remember $VIRTUAL before it's run to $5B? I'm here to $SERV a 🧵 about OpenServ⬇️
24
40
19
146
22,453