Pinned Tweet
i developed a language learning app for my wife and me (she's learning EN, im learning JA) and it's turned out nice so i'm opening it up to a small test group

rn focused on english, 中文, 日本语, and español

reply SMORT if you want in

iOS 26+ only, serious learners preferred
7
0
12
1,241
.@Sogni_Protocol is already doing decentralized inference and crushing it

base:0xe014d2a4da6e450f21b5050120d291e63c8940fd is undervalued
MEGA pivots from cursed L2 to a hyped AI category.

$MEGA could do a 10x from here at current $59m MC to catch up with other "AI trades"*:

By MC / FDV:

- $NEAR: $4.63b / $4.63b
- $TAO: $2.53b / $4.53b
- Pearl: $586m / $3.29b
- $VVV: $943m / $1.55b
- $GRASS: $342m / $473m
- solana:BZLbGTNCSFfoth2GYDtwr7e4imWzpR5jqcUuGEwr646K: $49m / $92m
- $ORBIO: $25m / $25m (stake to earn inference credits)

The problem is $MEGA's FDV :(

Only 10% of MEGA circulate and FDV is still at $513M.

But IF MEGA TEAM SHIPS, the token can moon. But we have time to see if the team delivers.

* I know I can't compare those tokens as they sell different AI products but for degens chasing AI narrative these tokens might come to mind.
0
0
1
194
coderdan (马卡龙) retweeted
What if Higgsfield was non-shady and had a token anyone can buy?

What if they’ve been quietly building for 3 years through the winter?

What if their revenue and monthly active users are both at record highs?

People say I don’t bullpost enough. They’re right.

I’d rather build and let the results speak for themselves.

And I believe what we’re building will be far bigger than any single DePIN network over time. Because we don’t just build the network we build the apps people want.

Bookmark this post for later.

Quiet momentum. Loud results.

base:0xe014d2a4da6e450f21b5050120d291e63c8940fd
3
3
14
599
no clue what i'm gonna do with it but it feels powerful
0
0
1
305
and if you want to learn Mandarin to improve your EV even further, use my app x.com/coderdannn/sta…
If you're a middle aged, single, balding (optional), normal western white man, the single most +EV thing you can do for yourself in this decade is going to China.

The odds are incredibly rigged in your favor compared to Chinese males.

The gap won't last forever though.
2
0
2
1,645
Serious language learners:

You've tried Duolingo but didn't want to feel infantilized
You've tried Anki but didn't know where to begin
You've tried Speak but progression is too slow

If this has been your experience, I welcome you to try Smort! It's not perfect yet by any oving rapidly every day.

Currently supports Spanish, English, Mandarin, and Japanese.

Interested? Comment SMORT + the language you want to learn.

iOS 26+ only.
1
0
2
390
Wow that was fast!

Maybe there really is alpha in adding native iPhone Duo support

Before i added support, my app was in the queue for 3d with no review

After adding support (took <1d), it got reviewed twice, both within 36hrs of submitting

thanks for the tip @RhysKentish
word on the street is that apple are prioritising app reviews that add duo support right now
1
0
2
465
7 years ago when i was building my first language learning app, I was worried that AI was going to disrupt the space and no one would want/need to learn foreign languages because AI made it unnecessary

it's still early in the AI revolution so that day still may come

but at least for now it's clear that demand for learning languages is still high

and now it's possible to use AI to 10x the speed at which we learn/teach languages

if you're interested in the intersection of AI and language learning, let's talk
4
0
10
483
Smort is an AI-powered learning app I developed to learn Japanese and help my wife learn English

It also supports Mandarin and Spanish

I'm gathering a small group of beta testers before it goes in the app store

Reply SMORT + the language you're learning if you're interested!
3
0
4
334
YAY i finally got my claude app to shown an unread badge!

that was WAY too hard
1
0
3
338
Opus 5.5 knocked it out of the park, insane update

not only the new video generation capabilities

but CODING is nearly flawless

and uses FAR fewer tokens than Astra or Sol

i've been slamming Opus all week and only at 35% usage

can't imagine what a monster Fable 5.5 will be
2
0
3
480
Opus 5.5 is undoing all of the overengineered bs that Astra and Sol added to my codebase

nature is healing
0
0
1
355
coderdan (马卡龙) retweeted
onchain frens 4ever 👻
7
11
56
4,288
China is leading the US in so many areas but it’s sad that they still don’t allow FSD, and iPhones bought in mainland don’t support Apple Intelligence

this is a real and growing QoL gap and makes me bearish on China until they fix it (hopefully soon, still love this country!)
4
0
3
749
i don't know who needs to hear this but if you have multiple codex subs, make sure you burn through all of the resets on a single account first before moving on to the next one.

don't be stupid like me and spread the reset usage out, it just screws up your entire reset time
0
0
3
491
ok i admit obsidian is pretty sick
2
0
6
752
negative prompts are more powerful than prompts
0
0
5
554
.@thsottiaux it would be neat if we could "mute" a thread in Codex so we don't get notifications for it

useful for threads running automations for those of us who have OCD and have to hit notification 0 on everything
0
0
3
508
omg did they really just reset again the madlads
0
0
2
402
I got tired of manually guessing which model was best for the job in Codex, or always choosing Sol 5.6 Max and wasting tokens, so I made a skill!

Introducing "auto-dispatch" for Codex

Invoke it with Sol 5.6 Max, tell it what to do, and then it will automatically route to the and archive the initial thread.

Works with /goal too!

PRs welcome 🤝
5
1
18
702
Replying to @coderdannn
Point your codex here to try it out:

github.com/cinnabarhorse/…
0
0
1
138