Nate
Region: US
Monday 28 September 2026 00:03:16 GMT
16459
597
31
30
Music
Download
Comments
The Bucky Verse :
89 million tokens. how did you do that. million tokens context in one session. Im confused
2026-09-29 20:17:12
0
Eddy :
Hmmm that’s pretty dang cool!
2026-09-30 05:24:32
0
Panda AI :
Where is the Lego set
2026-09-28 02:37:02
12
Toasted Rye Guy 🇺🇸🌎 :
So I have an 80’s set of legos. Thousands of random pieces. I took a picture of my Lego’s all spread out and asked ChatGPT, (don’t remover which model) maybe 3 months ago, to give me several build instructions for different things I could build with the random pieces. It failed. Just made shit up. Might be better today but that was my experience with legos. I’m sure it will eventually get there and may already be there. Haven’t tried with Astra.
2026-09-28 01:18:44
1
SandWorm666 :
I was wondering where this mythical .50 cent lego set is… 😂 Then I realized what you meant.
2026-09-28 03:29:10
4
mt_trout_bum :
My reset was Saturday evening, it is now Sunday evening and I have been working pretty hard on a product design and have started working on ai vision recognition training on a separate long running task. I am very impressed with the build quality so far and am only up to 4% weekly. With Fable I would have been at 15-20% easily with this amount of work. I am very impressed.
2026-09-28 02:40:07
0
#TonightsShowBirdWhistler :
Thank you! I love this information and your real world examples! Thank you, thank you thank you AI Nate
2026-09-28 16:11:12
0
CAD :
so we gonna see this Lego set? or will you but a sub for us
2026-09-28 07:01:26
1
Bill Zhang :
Lego photos or it didn’t happen
2026-09-28 18:09:02
2
TasteofAlps🇨🇭 :
why can't we use AI to solve housing crises by building Lego like construction?
2026-09-28 07:56:38
1
Chris DiMarco :
Opus 5.5 is the best model I’ve ever used, very task efficient
2026-09-29 00:18:08
0
Victoria :
I had an issue this weekend where opus wouldn’t follow my step by step instructions. It took me 6 coding sessions to find one that would read prompt artifacts and follow instructions
2026-09-28 09:10:05
0
Lamina Flow :
I am going to try it. I have bought 20 quid of credits on codex accidentally set the model to astra and it burnt through 80% of the credits in less that 15 prompts . I am still new to this but very disappointed
2026-09-28 00:27:21
0
Eddie :
that's amazing
2026-09-28 17:47:01
0
user9762258332577858437 :
opus 5.5 seems to cost much less overall...
2026-09-28 22:51:20
0
757 Virginia Beach Influencer :
Do you have a Lego Set Skill so I can take a picture of my room and turn it into a Lego set...
2026-09-28 22:10:29
0
DimLei :
What did you do with the result ? Did you order the parts ?
2026-09-28 18:05:20
0
DadPlaysGames 🇸🇪 :
Devin Fusion is worth talking about too. it's basically pairing Opus, Fable or Astra with its own SWE-2 model as a sidekick doing all the grunt work for the main model. Massive cost savings for me
2026-09-28 10:47:19
0
HolzMS-Graz :
🤣🤣🤣🤣 i got 300kg of lego and 50kg of technik. would cost a quadrizillion €
2026-09-28 12:11:32
0
Larry Bowden :
How did you accomplish that with subscription, with the chat web UI?
2026-09-28 00:22:47
0
Bricks in a Bag :
Nate, this is right up my street! How can we see this in working or the final results?!
2026-09-28 08:13:43
0
nordictravk :
The subscriptions (at least for Claude) are a massive subsidy on the API. I know it’s not direct math but if your 89 million tokens was 2%, then you could comfortably burn over a billion tokens without breaking a sweat on your subscription. It’s a bargain if you ask me
2026-09-28 04:41:48
0
Tanner Stone :
I had an auto tune on my cars in my video games
2026-09-28 03:05:42
0
Them Apples :
wooow can you share? this would help me with my project
2026-09-28 06:12:58
0
Coder Since '84 | Vibe + AI :
Cool idea 🥰
2026-09-28 00:13:01
0
To see more videos from user @nate.b.jones, please go to the Tikwm
homepage.
