← Back to leaderboard
aleximarkett

alexi

active
@aleximarkett

Trader. Builder. AI enthusiast. Investor in Anthropic, Palantir & 500+ more. Exploring the future, one bet at a time 🌎

169,064 followers · 1,825 following · Last polled 3h ago
Perf. Score
66.8
Rank #92
Value Score
—
set rates to compute
Median views
12K
p25 10K · trailing 7d
Followers
169K
+0.2% 7d
ER (impr.)
1.57%
eng ÷ impressions
ER (foll.)
0.15%
eng ÷ followers
Est. CPM
—
needs a rate
WoW views
+16.4%
flat

Rates & value · est. from 12K median organic views

Quote tweet
—
no CPM
Post
—
no CPM
Thread
—
no CPM
Retweet
—
no CPM
rate age unknown (pre-audit import)

Trends

Follower growth
Median views / day
Engagement rate
View distribution · trailing-7d posts

Each blue dot is one organic post. The Performance Score ranks by the median (12K), not the mean (16K) — so a single viral post doesn't inflate a creator's reach. The p25 floor (10K) is the level 75% of posts clear.

Recent posts · ranked by views

PostPostedViewsLikesRepostsRepliesQuotesBookmarksER
i think everyone is too focused on building bigger AI models while overlooking something equally important: making existing models run faster. that's what caught my attention about EAGLE-4. their latest paper reports 4x the token throughput of standard decoding in its benchmark setup, with strong results even when hundreds of requests are running at once. personally, i think getting more performance out of the hardware we already have is a much bigger deal than people realize. we need more improvements like this in AI, especially as more developers start building with agents. worth checking out: https://t.co/w3dQ2UtYx92d ago84K2863169010.46%
I’m bullish on AI agents, but this is the part that worries me. A phishing email can trick a human. Now it can also try to trick the AI working on their behalf. As agents get access to inboxes, files and financial tools, security can’t stop at the model. It has to include what the agent can actually do. Smarter agents need tighter permissions.frozen4d ago16K163
25
68
0
1
1.58%
I’ve been waiting for trading agents to move from demos into actual financial workflows. The interesting part to me isn’t that an AI can analyze a chart or suggest a trade. It’s that I can give an agent a strategy, let it run continuously, and have it execute when my conditions are met. That changes the game. Trading starts looking less like sitting in front of a screen all day and more like designing a system you trust enough to run without you.frozen23d ago16K1642268011.60%
I think we've been focusing too much on making AI models smarter while ignoring how they actually use computers. AI is getting better every day, but it still struggles with basic tasks that involve clicking through websites and getting work done. @PineAIAssistant is taking a different approach with Pine Computer. instead of building another model, they're building a computer specifically for AI, where it can actually complete tasks and give you finished work. I think this is an interesting direction because the future of AI isn't just about giving better answers. It's about getting real work done without us having to manage every step.frozen13d ago16K1632232021.41%
Could you live without your phone for $1M? I think I could. I’m taking the money & I think a lot of you would too. So I put it on @POVMarket : https://t.co/E49bvCpYNI It’s in the Daily Pot right now with increased incentives for a limited time.frozen28d ago15K1732465131.78%
I’ve been waiting for tokenization to move beyond simply putting individual assets onchain. This feels like a much bigger step. Taking an entire portfolio strategy and packaging it into one transferable token changes how I think about accessing diversified investments. If this works at scale, I can see portfolios becoming composable building blocks across wallets, exchanges and DeFi instead of something locked inside a traditional brokerage account. That’s where tokenization gets really interesting to me.frozen28d ago15K1572021011.35%
Creatify is getting seriously good at turning AI video into something that actually feels ad-ready. this Snapchat ad is probably the best example of why this caught my attention. the concept is creative, the branding stays consistent, and the whole thing feels intentional instead of looking like a random AI-generated clip. and this was made in one shot. Opus 5.5 + Boreal-H3 is insane. definitely worth trying inside Creatify’s Model Playground.frozen22d ago15K1542336001.45%
Damn this could be wild heard neymar launching on sol. Not sure the last time a star this big had a launch. Gonna ape when its live.8d ago15K1783575011.97%
I’ve always thought the weakest part of KYC isn’t the idea of proving who you are. It’s how much information gets collected and stored just to prove one simple fact. ZK proofs change that equation. If I can prove I’m eligible, over a certain age, or not on a sanctions list without handing over my name, address and entire identity, that feels like a much better model for digital finance. Crypto has the technology to make privacy-preserving compliance possible. Now the interesting question is whether regulation catches up.frozen27d ago15K1632341011.56%
I’ve been thinking about Ethereum less as a blockchain and more as a cryptographic computer. That shift matters. If things like lean consensus, better state management, formal verification, private mempools and stronger cryptographic primitives all come together, Ethereum starts looking less like a chain that simply processes transactions and more like a platform for building entirely new types of applications. I think that’s the part of Ethereum’s roadmap the market still underestimates. The boring infrastructure work today could end up defining what’s possible onchain tomorrow.frozen25d ago15K1712474011.85%
I see a lot of parallels between AI and trading. In both, people tend to underestimate how much of an edge comes from having the right process. AI gives me the ability to research faster, test more ideas, and automate the boring stuff. Trading teaches me that none of it matters if I can't manage risk or control my own decisions. I'm not trying to predict everything or automate every move. I just want to make better decisions, faster, without taking stupid risks. That's the kind of edge I'm trying to build.frozen10d ago15K1572862201.72%
The headline sounds bearish, but the chart tells a different story. Only 2.2% of US households are paying for AI right now which means adoption is still incredibly early. What stands out to me is the direction: paid AI subscriptions have gone from almost nothing to a steadily rising share of households in just a few years. If AI becomes as normal as paying for Netflix, Spotify, or cloud storage, the real market is still mostly ahead of us.frozen17d ago15K1462764011.64%
I’m pretty bullish on AI, but I don’t think the labs building it should be the only ones deciding where it goes. If these systems are going to shape how we work, build and live, I want independent standards that let people actually measure safety, compare failures and hold companies accountable. Competition matters too. The future of AI shouldn’t belong to whoever has the biggest compute budget.frozen1mo ago15K1541268011.62%
I’m starting to think the real AI advantage isn’t just having a smarter model. It’s having a workflow that remembers what you’re building and helps you keep moving. That’s why I’m interested in tools like Claude Code Projects. Less time rebuilding context from scratch. More time actually shipping things. The next productivity leap might come from making AI feel less like a chatbot and more like a workspace.frozen13d ago15K1652085131.89%
This is the kind of crypto adoption I actually care about. I don’t need another company saying it’s “exploring blockchain.” I want to see crypto quietly becoming part of products people already use. 82M Galaxy devices is a massive distribution channel. If sending USDC across borders becomes something people can do directly from a phone they already own, that’s when stablecoins start feeling less like a crypto product and more like financial infrastructure.frozen15d ago14K1542661111.68%
I’m interested in Jev because trading automation doesn’t always need a huge LLM. Sometimes I just need a fast probability check on a headline. Right model, right task. That’s where the real efficiency gains happen.frozen5d ago14K1692660141.80%
I’m bullish on AI, but I don’t think trust should be the foundation of deploying powerful agents. Control should be. If an AI agent can access sensitive data, move money, execute trades or make decisions on my behalf, I want strict permissions, complete logs and a kill switch that doesn’t depend on the model itself. I wouldn’t give a human unlimited access to critical systems just because they seem competent. Why should AI be any different? The goal isn’t to make AI less capable. It’s to make its capabilities usable without blindly trusting every action it takes. The smarter the model gets, the more important the guardrails become.frozen11d ago14K1732284521.98%
I think the biggest AI blind spot right now is that most people are judging the technology based on what they’ve personally experienced. If you’ve only used AI to summarize emails or answer questions, the hype probably looks ridiculous. But when you start using frontier models to build software, run complex workflows and solve problems that used to take weeks, your perspective changes completely. I’m fascinated by how differently people are experiencing the same technology. The gap between what AI can do and what most people think it can do might be one of the biggest opportunities of this decade.frozen10d ago14K1642863211.79%
I’m really interested in where AI interfaces are heading. I can already ask AI to analyze data, but turning that data into a live dashboard or an animated explainer without switching between five different tools is a different level of productivity. The less time I spend formatting information, the more time I can spend actually understanding it. AI isn’t just helping me think anymore. It’s starting to change how I turn ideas into things people can see and use.frozen14d ago14K1592565111.75%
Grok 4.7 taking the top 3 spots on VulcanBench is pretty impressive. What makes this more interesting is that this isn’t just a “write some code” benchmark. The models have to rebuild old software behavior, work with hidden tests, and handle real engineering constraints. That’s much closer to actual software work. For me, the big takeaway is that coding models are moving from generating snippets to handling full repository level tasks. That’s a much more meaningful jump.frozen16d ago14K1542460021.67%
Shayne teasing “a little something” at Token2049 right after Polymarket puts up numbers like this… are we getting closer to hearing about $Poly now?18d ago14K1353118021.30%
Pyth's Q3 Connect just wrapped up and all I can say is wow. I feel like the 100% Rule was easily the biggest announcement. 100% of the funds the Pyth DAO receives from its products will now be used to buy $PYTH on the open market. And this is happening while Pyth is already working with names like Nasdaq, Coinbase, Kalshi and Polymarket. Not to mention 94.1% of RWA perp volume in Q3 was on venues using Pyth for pricing. If they keep growing at this rate, those $PYTH purchases could start adding up pretty quickly. Surprised more people aren't talking about this tbh.frozen14d ago14K1852862011.93%
Trading today is a perfect reminder that the market doesn’t care how confident you are. You can have the cleanest setup, perfect entry, and still get stopped out in minutes. Then the market moves exactly where you expected, just without you. That’s trading. You don’t need to catch every move. You just need to manage your risk well enough to still be here for the next one.frozen20d ago14K1632463011.76%
I’m more interested in AI tools that help me make money, not just build things. Building the product is step one. Getting customers is the real game. AI is starting to connect building, marketing, and growth into one workflow. That’s where the real leverage is.frozen8d ago14K1722768011.90%
I’m watching AI interfaces more closely than model benchmarks lately. The real shift is from AI that tells me what to do to AI that can actually do the work. Imagine an AI that watches market conditions, researches a setup, tests the thesis, and brings me a trade plan with the risks clearly laid out. I still make the final call. But the research and execution become dramatically faster. That’s the interface I want: less prompting, less tab-switching, more useful decisions. The next big AI product might not be a better chatbot. It might be a better way to work.frozen8d ago14K1562365011.74%
I’ve started thinking about leverage very differently. In trading, leverage lets you control more capital with less capital. In AI, leverage lets one person do the work of what used to require an entire team. In crypto, leverage shows up everywhere too capital, liquidity, composability, distribution. But leverage cuts both ways. It can multiply your upside, but it also makes bad decisions more expensive. That’s why I think the biggest edge in both AI and trading isn’t having more leverage. It’s knowing when to use it.frozen23d ago14K1622465011.79%
I think this is a bigger move than it looks. Meta already has the distribution. Billions of users, millions of businesses, massive amounts of data and attention. Now they’re building the enterprise layer on top of it. If Meta can turn AI into something businesses use to sell, automate, support customers and actually run parts of the business, the opportunity is much bigger than just another AI assistant. The AI race is slowly becoming a distribution race too.frozen24d ago14K1652866121.87%
I've been thinking about what AI agents will actually change about trading. Finding information was never the hardest part for me. Knowing when to act, when to wait, and when to protect capital is. AI can automate the research, monitor setups and execute predefined rules. But if the strategy is bad, automation just helps me lose money faster. The edge isn't letting AI trade everything. It's knowing exactly what I should let it do and what I shouldn't.frozen11d ago14K1672677201.96%
The market feels like it’s entering that phase where risk appetite comes back before everyone fully believes it. Bitcoin is holding around $84K, while higher-beta alts are starting to move harder. What I’m watching now isn’t just price. I’m watching where the liquidity rotates next. That’s usually where the interesting trades start showing up.frozen26d ago14K1692472011.95%
Saturday hits different. No alarms. No deadlines. No pretending to care about emails. Just a fresh weekend, a little time to reset, clear the head, and maybe rethink a few decisions made during the week Markets can wait for a bit. Enjoy the weekend, recharge, and come back with a better mindset.frozen19d ago14K1511963011.72%
i think a lot of founders are still underestimating how much AI is changing content marketing. this is a pretty good example. almost 9M views on TikTok without filming a single video or hiring creators. what caught my attention is how they've automated almost the entire process. Viktor studies what content is already working, Claude and Higgsfield create the slideshows, and Postiz handles the scheduling. 120 slideshows created in one go, with AI also tracking results and figuring out what to post next. i've always felt that understanding what makes content work matters more than just producing more of it. and now that AI can handle most of the execution, i think we're going to see a lot more small teams competing with much bigger marketing budgets.1d ago14K1642757041.86%
Trading has made me realize that the hardest part isn’t finding opportunities. It’s knowing when not to take them. There will always be another narrative, another breakout, another coin that looks like it’s about to run. I’d rather miss a move than chase something just because everyone else is excited about it. The goal isn’t to trade more. It’s to stay liquid, stay patient, and be ready when the setup actually matters.frozen27d ago13K23148000.33%
I’m less interested in AI getting smarter and more interested in AI becoming genuinely useful. Give me an agent that saves me 3 hours every day over a model that wins another benchmark. Intelligence is impressive. Execution is the real edge.frozen7d ago13K1672363111.90%
I think AI becomes truly useful when it stops making me switch between tools and starts working inside the ones I already use. Reading tabs is useful. Taking action across them is the bigger shift. The browser might become the interface for AI agents.frozen7d ago13K1882161002.01%
The market is becoming less about picking the “best company” and more about understanding where capital is moving. AI is reshaping software. Crypto is reshaping finance. Robotics is reshaping labor. Energy is becoming a bigger bottleneck. The interesting part is that these trends are starting to overlap. I’m less interested in chasing whatever is pumping today and more interested in finding the companies sitting at the intersection of these shifts. That’s where I think the biggest opportunities of the next decade will come from. Follow the change. Then follow the money.frozen18d ago13K1502365011.78%
I’m excited about the boring infrastructure upgrades that make AI agents actually practical. If an agent can keep its context cached for hours instead of paying to rebuild it every time, persistent workflows become cheaper to run. The real AI race isn’t just smarter models. It’s making intelligence cheaper to use, continuously.2d ago13K1522760001.79%
Today’s market is a good reminder that a green candle doesn’t automatically mean the pressure is gone. BTC is trying to recover, but ETF outflows, liquidations and macro uncertainty are still making this market difficult to read. I’m not interested in catching every bounce. I want to see buyers follow through before I get confident with a position. In markets like this, protecting capital matters more to me than proving I called the bottom. I don’t need to catch the first move. I need to be there for the right one.frozen9d ago13K1582562011.85%
I’m paying more attention to how long AI can stay useful on complex tasks, not just how well it answers a prompt. Better coding is great, but persistence, long-context reasoning, and the ability to work through difficult problems are what really interest me. If models keep improving at that pace, the bottleneck shifts from what I can build to which problems are actually worth solving. That’s the part of this AI race I’m most excited about.frozen9d ago13K1591959011.80%
One thing trading and AI have taught me: More information doesn’t automatically give you an edge. Better decisions do. AI can scan thousands of data points, spot patterns and help me test ideas faster than ever. But it can’t magically turn a weak strategy into a good one. And in trading, acting on every signal is a great way to lose money. I want AI to help me think better, not just move faster. Because the real edge isn’t having the most data or the fastest tools. It’s knowing which signals actually matter.frozen12d ago13K1531457111.71%
I’m not seeing a market that needs me to force a trade right now. BTC is hovering around $83K–$84K, ETH is around $2.6K, while longer-term Treasury yields are putting pressure on risk assets. ETF inflows are still positive, but the pace has slowed. That’s usually when I get more selective. I’d rather watch liquidity, wait for confirmation, and miss a move than convince myself I need to be in one. Patience is a position too.frozen22d ago13K1562266011.85%
I feel like crypto spends way too much time celebrating what’s visible. Token launches. TVL. Price action. Narratives. Meanwhile, something like PeerDAS quietly changes what’s possible at the infrastructure level. Getting decentralized nodes to agree on data availability without every node needing the full data is a massive engineering achievement. I’m much more interested in these boring-looking breakthroughs than the next shiny narrative.frozen26d ago13K1731565031.95%
I use different AI agents for different things, but giving them my card details or passwords directly never felt right. That’s why Agent Vault from @Selfxyz makes sense to me. You set it up once, connect it to agents like Claude Code, Codex or Grok, and they can make payments or log in without ever seeing your actual card number or password. You still control the limits, approvals and even have a kill switch. For me, this feels like the kind of security layer AI agents need before we trust them with more real-world tasks.frozen7d ago13K1832562112.08%
I’m generally more worried about what happens when we make inaction the default. Every new technology comes with hypothetical risks. That’s necessary to think about. But there’s an opportunity cost to treating every possible downside as a reason not to build. I see the same thing in AI, crypto and markets. You can spend forever modeling what could go wrong. At some point, you have to ask what we lose by never trying. Good risk management shouldn’t mean eliminating risk. It should mean understanding it well enough to keep building.frozen21d ago13K1592264011.90%
I keep coming back to one idea: AI is making execution cheaper, while trading is teaching people that execution isn’t the edge. You can build faster. You can analyze more data. You can automate entire workflows. But the real edge is still knowing what to build, what to trade, when to act, and when to do absolutely nothing. That overlap between AI and trading is what I find most interesting. Both are becoming less about doing more and more about making better decisions.frozen24d ago13K1722465112.09%
I feel like most people are still using AI for marketing in the most basic way possible. Writing captions, generating random ideas, and calling it a day. Came across something interesting where Pedro gave Muse, an AI agent he uses through WhatsApp, almost his entire app marketing workflow for 24 hours. It researched what was already working on TikTok, came up with slideshow ideas, created different versions, and even organized the posting schedule. The result? 300K+ views from the slideshows. What I actually liked is that Pedro still had the final say on what went live. That's how I think AI should be used. Let it handle the repetitive work while you focus on making the right decisions.2d ago12K1622858032.03%
Might've found my favourite meme on Base $EMOO by @emoorawrXD is leading the @feeldotcash stock race. Ton of upside here imo. Just send it already lmao https://t.co/bF0vrN54ds2d ago12K1683057022.13%
AI is making execution cheaper by the day. The interesting part is what happens next. When building becomes almost effortless, the edge shifts from “Can you build it?” to “Do you know what’s worth building?” Taste, judgment, distribution and timing start mattering a lot more. That’s a pretty exciting shift for builders.frozen29d ago12K1632069022.12%
I think this is where @polsia gets really interesting. The founder needs engineers, asks for contractors, and gets options back within minutes. Then it’s the same thing for branding, social, and other work. What I like is that it’s not just AI trying to do everything itself. When a human is needed, it can help find the right person too. For a solo founder, that can save a crazy amount of time that normally goes into hiring and managing people. That’s where I think Polsia becomes really useful.frozen29d ago11K1562133021.88%
been looking at some of the recent solana:HZ1JovNiVvGrGNiiYvEozEVgZ58xaU3RKwX8eACQBCt3 wallet activity and there's definitely something interesting going on a few larger wallets have been steadily building positions, and the timing is what caught my attention Pyth Pro has been growing insanely fast, going from around $1M ARR in late 2025 to $10.4M in August, and I've also heard there could be some pretty interesting news around solana:HZ1JovNiVvGrGNiiYvEozEVgZ58xaU3RKwX8eACQBCt3 value accrual coming at Pyth Q3 Connect obviously no idea what these wallets know, if anything, but seeing bigger holders accumulate while the business is growing like this definitely has my attention very curious to see what gets announced at Q3 Connect: https://t.co/Y0epSEZcUffrozen21d ago11K1662961012.30%
BTC dipping below $80K and bouncing back is a good reminder of how quickly sentiment can flip. With inflation still running hot and AI stocks under pressure, I’m not rushing to call a bottom. I’d rather wait for confirmation than turn volatility into an expensive lesson. My job isn’t to predict every move. It’s to manage risk when I’m wrong.frozen6d ago11K1712059012.27%
I love the idea of having an AI salesperson that never gets tired of posting Give it a website, let it learn the business, and suddenly content creation becomes a lot less painful. Now the real test is whether those 30 posts can actually bring in customers.frozen5d ago11K1622460012.24%
I think @WireNetwork is taking a much better approach than the usual “raise first, figure out liquidity later” model. their LCO is live, and what stands out to me is that 100% of the proceeds are going directly into network liquidity and nothing goes toward company operations. that matters because liquidity is one of those things every network needs from day one, but very few actually build around it this early. Wire is also working on cross-chain infrastructure where apps and assets can move across supported chains without relying on traditional bridges. If they execute on that, starting mainnet with deeper liquidity already in place makes a lot of sense to me. LCO hub: https://t.co/JIoErzQOt6frozen23d ago11K1642468012.34%
I’ve been waiting to see more AI companies think beyond intelligence, valuations, and growth. AI could accelerate scientific discovery, transform healthcare, and create enormous wealth. The question is how much of that progress actually benefits everyone. Anthropic committing roughly 20% of its shares to public benefit is a meaningful step. Building superintelligence is one challenge. Making sure humanity benefits from it is another.frozen4d ago9.5K1652962002.70%
I think Anthropic can easily launch above a $2T market cap. The demand around AI right now is just too big to ignore. I am using @Outcome_xyz for my prediction since they have the best price on the market right now. and even if my prediction doesn’t hit, there’s still a chance to earn part of the $1,000,000 rewards. w-w if you ask me.frozen17d ago9.0K1472337012.31%
Building an AI startup to $20M run rate in 340 days is crazy, but what I find interesting is how much happened behind that growth. From dealing with bugs and high AI costs to lowering prices and finding customers, it wasn’t just about building a good product. I think that’s the part people often miss. Building is one thing, but learning how to grow and keep going when things get messy is a whole different challenge. Impressive journey from @polsia.frozen3d ago8.5K1722558043.06%
I've always felt like AI is mostly something we use on our own. But most things I do involve other people, whether it's planning a trip with friends, figuring out dates, or just organizing something together. That's what caught my attention about @wabi. Instead of everyone using AI separately, you can actually use it together in one group chat. I feel like this is where AI should be heading. Less talking to bots alone and more getting things done together.frozen3d ago6.9K170926113.01%
I think the next big shift in AI is moving from assistants I have to prompt to agents that know what needs my attention. I don’t want to manage every task. I want AI to handle the work while keeping me in control of the decisions that matter. Less prompting. More progress, even when I step away.7h ago6.8K1303235012.90%