Software Engineer Helping you get 10x more out of AI Daily prompts, workflows & model tests

Joined July 2026
I asked Opus 5.5 to visualize the biggest scam in humanity. No words allowed. It gave me this.
71
85
10
2,659
439,198
Ivana retweeted
First year of computer science, I thought hardware just wasn't for me. Then I built my first circuit and watched an LED blink because of code I wrote. Best feeling ever. That's when it clicked. Hardware gets way easier when you break it down into small steps. So I built a free Raspberry Pi lab where you can wire circuits, write Python, and watch it all come to life. Got a hardware exam or project coming up? Start here 👇 ivanainai.com/pi-lab
11
2
34
1,810
Ivana retweeted
OpenAI Codex, Day 21 of 28: HALF the shipping streak has been resets. 💀 9 full resets. 1 banked reset. 20 days. The ~50% faster models, persistent chat caching, and Day 19 cache fix? Those were genuinely useful. But calling quota refills “shipping” is doing some serious heavy lifting here. Tibo says “fun week ahead.” Seven days left. Prove it.
5
1
19
2,238
First year of computer science, I thought hardware just wasn't for me. Then I built my first circuit and watched an LED blink because of code I wrote. Best feeling ever. That's when it clicked. Hardware gets way easier when you break it down into small steps. So I built a free Raspberry Pi lab where you can wire circuits, write Python, and watch it all come to life. Got a hardware exam or project coming up? Start here 👇 ivanainai.com/pi-lab
11
2
34
1,810
OpenAI Codex, Day 21 of 28: HALF the shipping streak has been resets. 💀 9 full resets. 1 banked reset. 20 days. The ~50% faster models, persistent chat caching, and Day 19 cache fix? Those were genuinely useful. But calling quota refills “shipping” is doing some serious heavy lifting here. Tibo says “fun week ahead.” Seven days left. Prove it.
5
1
19
2,238
Ivana retweeted
Grok Bot can make phone calls for you now🤯 It dials, sits on hold, presses the menu keys, gets transferred, and sends you notes when it's done. Here's where it actually saves you hours 👇
1
2
11
737
Grok Bot can make phone calls for you now🤯 It dials, sits on hold, presses the menu keys, gets transferred, and sends you notes when it's done. Here's where it actually saves you hours 👇
1
2
11
737
5. The one call you've been avoiding for 3 weeks. You know the one.
1
38
Tip: give it everything up front, like your account number, what you want and your deadline. It can only ask for what you told it.
35
Ivana retweeted
Unpopular opinion: your results with AI depend heavily on how smart you are.
359
413
78
4,477
155,109
Unpopular opinion: your results with AI depend heavily on how smart you are.
359
413
78
4,477
155,109
Ivana retweeted
Cheap Model of the Week: NVIDIA Nemotron 3 Ultra. I deliberately wrecked a JavaScript physics simulator. NVIDIA's FREE model fixed the entire mess on its first try. Gravity worked backward. Pause was decorative. Reset turned 240 particles into 720. Nemotron fixed all three without wrecking the UI. $0 spent. Watch the before and after. Broken code + exact prompt in the replies so you can test it yourself.
8
5
17
830
Cheap Model of the Week: NVIDIA Nemotron 3 Ultra. I deliberately wrecked a JavaScript physics simulator. NVIDIA's FREE model fixed the entire mess on its first try. Gravity worked backward. Pause was decorative. Reset turned 240 particles into 720. Nemotron fixed all three without wrecking the UI. $0 spent. Watch the before and after. Broken code + exact prompt in the replies so you can test it yourself.
8
5
17
830
Here's the broken JavaScript simulator + exact repair prompt: github.com/emberlumo/redpen/…
2
162
Ivana retweeted
How much are we overpaying for AI coding out of habit? I've been testing Kimi K3.1 Pro against Opus 5.5 and Astra 6 since yesterday. Interactive games, frontend logic, 3D builds. Kimi beat both on some of my tests. It's not flawless. I'd still trust Opus with longer, more complex coding sessions. But if Kimi can deliver working code at half Opus's price and a fifth of Astra's, why am I reaching for the expensive model first?
8
1
14
1,333
How much are we overpaying for AI coding out of habit? I've been testing Kimi K3.1 Pro against Opus 5.5 and Astra 6 since yesterday. Interactive games, frontend logic, 3D builds. Kimi beat both on some of my tests. It's not flawless. I'd still trust Opus with longer, more complex coding sessions. But if Kimi can deliver working code at half Opus's price and a fifth of Astra's, why am I reaching for the expensive model first?
8
1
14
1,333
Ivana retweeted
"Everyone's leaving Anthropic" is two people. Thariq and Anthony Morris, both Claude Code, both gone in a day. Real loss. Not an exodus. Calm down.
7
4
1
14
889