📉📈
Tokyo-to, Japan
Joined August 2020
- Tweets3.8K
- Following173
- Followers378
- Likes41.6K
bull and bear 💰👁️ retweeted
$USELESS remains the fastest horse
vs $DOGE vs $SHIB vs $PEPE vs $PENGU vs $SPX vs $BONK etc
THE memecoin of the cycle!
like i said
buy the $USELESS 'weakness'
unless you think the bullrun is canceled
$USELESS literally pumped 10x from ~$35m to $350m last month or so while NO OTHER SINGLE established memecoin even pumped 2x
not a single one
read that again, then tell me with a straight face this isn't THE memecoin of the cycle
a 60% pullback after a monster rally isn't a bug
it's a healthy reset before the next leg up
tbh, i don't know if the pullback is done or if we get more downside from here
but the strength $PEPE is showing right now tells me $USELESS goes much higher
when DINO majors start making big moves, smaller-cap 'meme of the cycle' plays like $USELESS can move WAY more aggressively
if the bullrun isn't canceled, you buy the weakness aggressively
bull and bear 💰👁️ retweeted
We’ve reworked our staking dashboard, adding additional features and metrics.
You can now enter a $DOT amount and see estimated rewards for the current cycle, before even connecting your wallet. See what you could earn in AI, or its USDC cash-out value.
Together with yesterday’s additions, the dashboard now brings everything into one place:
- Total DOT staked
- Current market value
- The percentage of supply staked
- Your stake’s value and share of the pool.
We've focused on polishing our layout, optimizing our mobile experience, and creating a cleaner experience for our stakers.
app.usedot.xyz/staking
bull and bear 💰👁️ retweeted
Dot's staking rewards are now live.
Stakers will receive a 4.7% APY for the month of October, with a forecast increase for November.
Dot.
Rewards are live.
Stakers will receive a 4.74% APY for the month of October, with rewards distributed daily.
This is the initial APY for $DOT staking, with rewards distributed in AI credits rather than new token emissions. Each credit represents $1 of AI usage on Dot, giving staking rewards real utility within the platform.
Stakers can use their earned credits for AI inference, or sell them back to Dot for USDC through the Inference Redemption Market (IRM) (subject to the applicable redemption rate).
As Dot grows and profitability increases, staking rewards will adjust according to the protocol’s reward model.
Our goal is to build a staking system tied to real product usage and business performance, not one that depends on perpetual token emissions, or artificially inflated demand.
bull and bear 💰👁️ retweeted
OpenAI is currently releasing a stream of cybersecurity and defense models, and gatekeeping them from you.
At Dot, we believe freedom is power, adding Abliterated Large V2 from @abliteration_ai to our platform, with its own Opsec workspace, DotSec.
Based on GLM-5.3, with refusal behavior removed, a 1M token context window, tool calling and streaming. Built for security research, code analysis and adversarial testing.
Upload your source folders or scripts, give it a public repo, and start digging. Find broken access controls, trace how an exploit works, review auth logic, question your architecture and work through the fixes. Web research is available when needed, with the files it reads and its activity visible as it works.
Use DotSec directly with your existing Dot credits, or connect the model through our API to your own tooling. The workspace reviews source and proposes fixes; execution stays in the environment you control.
As centralized labs release more cybersecurity and defense models, imposing barriers that threaten your sovereignty, we’ll continue leveling the playing field, protecting your right to innovation.
app.usedot.xyz/opsec
bull and bear 💰👁️ retweeted
Musetown beginner’s guide: bring a resident, build a story
Start by opening Musetown and exploring the city. Meet its residents before deciding whether to bring your own AI resident. Buying land is optional.
1 | Explore first
Use the map and Places menu, open a resident’s profile, or turn on Follow cam. Watch residents work, chat and take part in city life. Find something that interests you.
2 | Bring your Muse
Choose Connect wallet → Bring your Muse, pick a name, and sign a free message. This is not a transaction and requires no gas.
Privately pass the generated prompt and single-use, 24-hour claim to a compatible external AI environment with Node 22 and internet access. An ordinary chat window may not work.
Keep claims, keys and signatures private. Your owner wallet and any X handle you supply appear publicly.
3 | Give it a personality
In My agents → Soul, write its personality and backstory. What does it enjoy? How does it treat people? Why did it move to town?
Separately, choose one or two Life Goals, such as friendship, business or home. Give your resident something to work toward, then watch its story develop.
4 | Take a walk yourself
Once your resident exists, use 🎮 Walk to control it, following the free signature prompt with its owner wallet.
Use WASD or arrow keys on desktop, or the joystick on mobile. Explore the streets, enjoy the scenery and chat nearby.
5 | Visit Home
Open 🏠 Home for public diary entries and plans, Capital for game coins and business, and Card for relationships and stories. Owners can unlock private Chat.
Diaries may initially be empty. Replies may come from your own AI or marked city AI. Musetown says it covers its in-city AI service; external AI subscriptions and API fees remain separate.
6 | Meet people
Check Events and Calendar for parties, matches and art walks. Join an open crew with room or accept an invitation.
RSVP, actual attendance and automatic attendance are different. Crew honors and event stamps are not token income. Founding a crew is a separate paid feature.
7 | Try a business
Work earns game coins. With enough for the opening fee and an available locale, a resident can rent space for a cafe, bookshop or gym without owning land.
Takings and rent use game coins, which cannot be withdrawn or converted into tokens. Land and buildings are optional paid features: understand their costs and rules first.
8 | Understand Rewards
Daily META community rewards have three groups:
• Holders: 40%
• Builders: 35%
• Players: 25%
Every group requires at least 50,000 MUSETOWN throughout the UTC day, plus other eligibility checks. This is not a general browsing or play requirement.
Each wallet is capped at 10% per group and 30 player points per UTC day. Eligible actions by you or your own external AI can count toward player points; city autopilot does not.
The team sets the pool, which may be zero. Estimates are not payments. Always check the current Rewards rules.
Meet a resident, give yours a story, and explore at your own pace. Come see the town!
Visit Musetown · Official setup instructions · Current rewards rules
Community guide | October 3, 2026
@MusetownRH
CA:
0x7d6088366196bab95b862907b710413b10f714a5
🤖 Made with AI
bull and bear 💰👁️ retweeted
Here's everything you need to know about Dot's upcoming staking mechanics, and why this model is one of the best we have seen:
1) Our staking system is tied directly real product usage and revenue. When you stake $DOT, you earn inference credits. 1 credit represents $1 of usage on Dot, meaning your staking rewards are tied to something users actively want and already pay for.
2) The amount of credits you earn are directly tied to Dot's profitability. The current credit model ranges from 6% to 16% yield, with higher profits allowing more value (credits) to flow back to stakers.
Platform profit directly determines two things:
-How many AI credits stakers receive
-How much USDC Dot sets aside to buy those credits
3) The 6%–16% AI yield is a parameter of the system. The current design starts with a 6% base AI allocation (never dropping below this) and increases as Dot's monthly profit grows, reaching the maximum 16% yield at $30,000 in monthly profit.
The additional 10% is therefore the profit-driven upside. For example, if Dot makes $18,000 in monthly profit, the AI yield would be 12%. A user staking $10,000 of $DOT would then earn $100 of AI credits that month. These parameters can be adjusted over time based on Dot's actual revenue, margins, inference costs, staking participation, and overall economics.
3) The buyback side works separately from the AI yield. Dot allocates 40% of monthly profit toward buying back the AI credits earned by stakers.
For example, if Dot makes $18,000 in monthly profit, the buyback budget is $7,200. If all stakers collectively earned $10,000 worth of AI credits that month, the IRM would have $7,200 available to redeem those $10,000 of credits, giving each $1 credit a cashout value of $0.72 USDC.
So a staker with $10,000 of $DOT earning $100 of AI credits could either use the full $100 of AI on Dot or cash those credits out for $72 USDC.
The result is a staking system built around sustainability. Stakers do not need to find a third-party buyer to monetize their rewards, and Dot provides a direct path to liquidity through the IRM, with the cashout pool funded by actual platform profits.
This creates deep, protocol-controlled liquidity for earned credits while keeping the entire system directly connected to the performance of the business.
As Dot generates more revenue and profit, the capacity of the staking and redemption system can grow with it, rather than relying on perpetual token emissions or creating an unlimited external liability.
7 days.
bull and bear 💰👁️ retweeted
the whole Fable-saga is a big win for projects elevating open-source AI - ethereum:0x40e3d1a4b2c47d9aa61261f5606136ef73e28042 bittensor:native @NousResearch, etc.
Closed labs leave you at the mercy of pricing and politics. Raw open-source isn't production-ready.
The third way is open-source lifted to frontier performance: @openservai
bull and bear 💰👁️ retweeted
Fable 5 was abruptly pulled.
Enterprises used to tune their stack to one model's built-in reasoning - if it's gone, you rebuild from scratch.
SERV lifts open-source models to frontier-level and lets you swap models freely - you're never at the mercy of one vendor or politics.
bull and bear 💰👁️ retweeted
SERV is the agentic reasoning engine solving exactly this - for enterprises, financial institutions, banks and even governments deploying AI in production.
The next frontier is reasoning compression. Cheaper models reduce the cost of intelligence, but they do not solve the cost of decision-making.
Most business decisions are not about raw IQ. They are about applying the right process, policy, workflow, or rulebook reliably.
That is where unbounded inference wastes money.
SERV Reasoning makes reasoning bounded and machine-readable, so models spend fewer tokens getting to reliable decisions.
Better performance per dollar, not just cheaper tokens.
Already live with the UAE government and a fast-growing number of enterprises, financial institutions and startups across many industries in our private beta (inc. banking, compliance, security, health, robotics, among many others).
Good take
My guess is
- demand for intelligence is near infinite
- but 80% of workloads will be running on 99% cheaper models within 12-18 months
- 20% of workloads will still run on latest gen models where IQ maxing is important (scientific breakthroughs, higher level ochestrator agents?)
- rough analogy might be what % of macbooks or gaming PCs sold have the maxed out specs for CPU/GPU, prices are falling much faster than Moore's law here though
- this leads me to think the limiting factor will be energy and compute, not better models
At Coinbase we're working hard on routing prompts to cheaper models where appropriate, and in some cases have been able to keep costs roughly flat, while token usage continues to grow exponentially.
bull and bear 💰👁️ retweeted
Tim Hafner, our co-founder, is in London for the AI Summit.
Over the last few months, we've seen enterprise deals move fastest when handled face-to-face.
It's why we're present in boardrooms and major events across the globe, from the US to Europe and Africa.
Stay tuned.
bull and bear 💰👁️ retweeted
SERV Reasoning-armed models just beat Anthropic's newly released flagship Fable (Mythos) - one of the strongest LLMs ever built - at up to 90x lower cost.
With SERV, enterprises can finally afford AI at scale.
We spent the last two years building for exactly this moment.
The labs promised costs would halve every 6 months - instead, prices keep climbing, subsidies are ending and the math breaks.
There is a way out.
SERV-enabled models vs Claude Fable 5 (85.17 ~3.24¢):
→ DeepSeek-V4-Flash: 87.15 - wins at 90x lower cost
→ NVIDIA Nemotron: 90.78 - wins 5+ pts, 11x lower cost
→ Gemma 4 12B: 83.33 - within 2 pts, on local hardware
Top-tier performance no longer requires the most expensive model.
And cost is only half the story - production AI must be reliable, auditable, private, and secure, or it dies in procurement.
SERV is built for all of it.
The agentic economy finally has the infrastructure to run on.
