Alex retweeted
THE BEST WAYS TO MARKET YOUR MOBILE APP
---
TikTok Ads: Legendary, takes some time to figure out and your app needs to have good user economics but once you find winning creatives it's amazing. Also: TIKTOK ADS MCP
Meta Ads: Same, I prefer TikTok but it depends on your niche
Canvas UGC: The only form of UGC that's reliable. Posting on new accounts makes it both in-line with the nature of UGC as well as making it accessible to creators. Result? you pay little. It requires good understanding of UGC, good management skills, and good payment terms. Also, outreach can be tough sometimes
TikTok Farm: Usually with slideshows, which are only effective on TikTok. It does require many phones obviously, and good software to manage them. For software I recommend Ghost Farm by @wbyanme (not sponsored)
Organic slideshows: Highest ROI, organic way to post content for free. My go-to for testing the waters for a new app. Only on TikTok, the algorithm loves slideshows for some reason. Instagram not at all.
Reddit: Somewhat niche but works if you have a fitting audience. You can also artificially acquire upvotes, and AI indexes reddit too so win-win. Creates a lasting reputation for your app
X: You can go mega viral like with @nestymee if you have something good to show
Organic UGC: Good but you need to have acting skills and a good phone, and you need to understand the principles of UGC very well (appear professionally unprofessional)
Regular UGC: Inherently inferior to Canvas UGC. Why would you buy 1 video for $100? Only use case is if you need a quick video for a launch or website
ASO: Not a lifechanger but you need to have this done properly since it amplifies all your other marketing channels when people visit your app store page
Clipping (slideshows): Only works if your app has something visual to sell. Middle-man parties usually take a large chunk of revenue like 20%. Also suffers from some botting scams
YouTube: Can work, especially with automation like the legends from Prayer Lock did it. But Youtube Shorts has a much more immature audience than other socials
AI indexing: So underrated, so many people use AI as a search engine nowadays. The difference with AI vs regular search engine is that people kinda interpret AI as "good advice". So if AI recommends a product, it has social proof albeit artifically
Affiliates: Extremely niche, so far I only know 1 person who does it extremely well. He combined influencer with affiliate partnerships. There is no good affiliation platform that tracks attirbution without an affiliate code unfortunately. Consequently your affiliates get paid very little and they quit
SEO: Less about virality, much more steady but can be highly competitive. It's something good to have but not a gamechanger by itself just like ASO
ASA: App Store Ads are for most people only profitable with the free $100 ad spend by apple. Furthermore hard to scale
Influencers: Extremely niche. It can work but audience saturation is a huge problem. My first influencer deal was a banger but the videos afterwards sucked because his whole audience already knew us. Also most influencers are dicks, they want ridiculous amlunt of money and overvalue their work
Pinterest: If you have an aesthetic tumblr-like app it can work. I do think Pinterest Ads are un untapped channel
Snapchat: Very young audience and still unexplored territory for most of us. For Snapchat Ads: they match your ad spend well, so maybe worth a try
Substack: Garbage except when you target boomers, even then it's hard
E-mail: Do not scrape people's emails and spam marketing. You will only make people angry. Email is useful for re-subscribing users or abandoned transactions, or preserving a community but not as your main marketing channel
Product Hunt: Absolutely useless and should be banned for wasting entrepeneurs time. You can only win if you have a large following and even then, the people who read the newsletters are rarely your ICP. Garbage
Alex retweeted
Today, on Google Play we announced new ways we are evolving the Play Store for an experience that is unique to you and the content you enjoy most:
✨ Use Play to wirelessly install and run your own custom, vibe-coded apps privately on your device.
🔍 Discover apps effortlessly with Ask Play & Gemini
🎮 Dial in discovery with the all-new Game Finder & Gamer DNA
Alex retweeted
40 free sites crypto whales don't want you to know about
who's buying, who's dumping, who's about to unlock. all free
whale & wallet tracking
1.) arkm.com → see who owns any wallet
2.) app.nansen.ai → where smart money is moving
3.) debank.com → any EVM wallet's full portfolio
4.) app.cielo.finance → wallet alerts across chains
5.) whale-alert.io → every huge transfer, live
6.) lookonchain.com → whale moves explained in plain words
7.) kolscan.io → every KOL memecoin trade, realtime
8.) fomo.family → what top traders buy and why
9.) gmgn.ai → smart money and KOL tags on any solana coin
10.) solscan.io → solana explorer
11.) etherscan.io → ethereum explorer
before you buy a memecoin
12.) rugcheck.xyz → insiders, LP and dev wallet on one page
13.) v2.bubblemaps.io → linked wallet clusters
14.) insightx.network → hidden connections between holders
15.) solsniffer.com → solana scam score
16.) tokensniffer.com → EVM scam score
17.) dexscreener.com → top traders, KOLs and holders tabs
18.) birdeye.so → solana charts and wallet PnL
19.) geckoterminal.com → any chain, any pool
unlocks & insiders
20.) tokenomist.ai → unlock calendar
21.) defillama.com/unlocks → who gets tokens and when
22.) cryptorank.io → funding rounds and VC buy prices
23.) dropstab.com → vesting and investor lists
leverage & liquidations
24.) coinglass.com/hyperliquid → hyperliquid whale positions, live
25.) app.hyperliquid.xyz/leaderbo… → top perp traders
26.) coinglass.com liquidation heatmap → where the stops sit
27.) velo.xyz → funding, open interest, basis
28.) coinank.com → order flow and long/short ratio
where the money actually goes
29.) defillama.com → TVL, fees, revenue
30.) tokenterminal.com → protocol earnings like a stock
31.) cryptofees.info → who earns the most fees today
32.) app.artemis.xyz → money flows between chains
33.) farside.co.uk → daily ETF flows
34.) dune.com → thousands of free dashboards
35.) messari.io → free research
36.) l2beat.com → L2 risk ratings
37.) ultrasound.money → ETH burn and supply
38.) mempool.space → bitcoin fees and blocks
portfolio
39.) app.step.finance → solana portfolio in one place
40.) debank.com/stream → what wallets you follow are doing
bookmark this. you'll need it next time you're about to ape
Alex retweeted
We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks:
Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better:
Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better:
Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better:
Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work!
In summary:
- As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding.
- Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.
Alex retweeted
Stop what you’re doing and build a plugin extension for ChatGPT.
We’ve grown over 2000% since the announcement!
Alex retweeted
6.1 Sol at 1/5th of the price of Astra, and 95% cache read discount! Extremely capable model.
Alex retweeted
this guy might be a fucking genius
he built a fitness page around an AI generated blue character, posts simple workout / muscle growth slideshows and then sells a $19.99 workout blueprint through the page
no face
no filming workouts
no personal brand needed
just:
-> AI character
-> useful fitness content
-> consistent visual style
-> link in bio to the product
the page is already at 325k+ followers
he basically turned one made-up character into the face of an entire fitness brand
this is the kind of AI content use case i find way more interesting than random AI slop
Alex retweeted
Okay, everybody is doing it, but I couldn't resist: Here is Opus 5.5 one-shot of the prompt: "make a dynamic 15-second motion graphics video that shows what an incredible motion designer you are. Highlight the recent changes we made with the last few updates."
Probably needs more time because we are shipping so many things 👀
Alex retweeted
It took years to get the strategy and the indicator right.
Version 1, six years ago, was a single feed and a gut feeling.
What it does now:
- full aggregation across spot, futures and inverse, with each one separable
- detection of new positions being opened
- adaptive filters for downside and upside aggression
- single exchanges shown inside the full aggregation, so discrepancies stand out
- defender clusters and levels where aggression is strongest
Same idea as day one. Six years of tape to sharpen it.
Aggression into the highs is usually the last buyer.
Speed of Waves catches the moment the tape speeds up, which exchange it's coming from, and how much size is really defending the level underneath.
A fast tape, thin defense, and a level it has to come back to. So who's left to buy it?
New MMT labs video from @WhenSailing
Alex retweeted
this is how i ACTUALLY set up phones before posting to a US audience on tiktok 🇺🇸
1. get a cheap used iphone 11, it's usually around $50, and factory reset it
2. set the phone up in region United States
3. skip the Apple ID for now take the SIM out and turn off location, bluetooth, tracking, Siri + Screen Time
4. get a residential US proxy (New York / Miami reigon usually works) IPRoyal is the best place
5. Run the up theough Shadowrocket on iphone, it's like $3 and it’s easily the cleanest option
6. make sure the proxy is connected BEFORE you install anything
7. open Safari and make a fresh email then use that email to create a new Apple ID
8. go to:
Settings → General → Language & Region → United States
then:
Settings → General → Date & Time → turn off “Set Automatically” → choose a US city
9. download TikTok and sign up with the fresh email
10. quick check before finishing signup:
tap “phone number” first
if TikTok automatically shows +1, that’s the check we use to see if the setup is reading as US ✅
then switch back to email and finish creating the account
11. DO NOT post immediately warm the account up for around 4 days first
EVERYONE should run this setup when aiming for US audience
Alex retweeted
GPT-Live-1 is now available in the API.
Bring ChatGPT’s natural back-and-forth to your app, with voice agents that listen while they speak and work with the models and harness you choose.
Alex retweeted
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
Alex retweeted
0x7a41..c093
this wallet bought 41,000 memecoins on robinhood chain in nine weeks.
$1,200 in, $214,000 out
it did no research on any of them
94% of those buys went to zero. that is not the flaw in the strategy, that is the
strategy
it only works here because robinhood has been paying the gas since july 1. every
transaction is free, so a bot can do the one thing that's impossible everywhere
else: buy everything
22,600 new tokens launched in a single day at the peak. this wallet was in a large
chunk of them inside the first minute
i gave fable 5 all 41,000 buys and asked what it filters for. the answer was:
almost nothing
- one buy per token, four dollars, inside the first 40 seconds. never adds, never
averages, never comes back
- sells on a fixed multiple, not a story. off at 4x, doesn't look at what happens next
- nothing held overnight that hasn't already paid for itself
- the only filter is that the contract isn't a honeypot
roughly 300 buys out of 41,000 are the entire account. everything else is dust
this isn't skill. it's a lottery at industrial scale, played by a machine that
isn't paying for tickets
at four dollars a shot, two cents of gas per transaction turns 41,000 buys into an
$820 hole before a single token moves
the subsidy expires september 29
on the 30th this wallet either stops, or it was never about the free transactions


































