designing interfaces, testing agents, building strange little tools
- Tweets 3,374
- Following 281
- Followers 420
- Likes 1,251
Pinned Tweet
How are YOU here, this very morning?
It’s a fascinating question that’s dfficult to visualize, and I’ve attempted to have Opus 5.5 show just that.
2
5
339
Have been running errands and having coffee with friends
Where’s devday livestream and has it been cool?
3
1
149
ChatGPT appears to be quantized or nerfed or whatever the word is lol
Have been chatting with voice this morning to find some stats and some of its words are not even said correctly. It also has been contradicting itself a ton
Haven’t seen it this bad off since the free version of 4o mini
7
12
696
Here’s how I foresee the future
Either the frontier labs have a major breakthrough on efficiency and intelligence becomes cheaper
Or
Those who can afford the good ones will be fewer while the masses use cheap or open models, even with lower quality
3
2
100
If frontier labs are not making enough to keep themselves going, how can they NOT raise costs?
With OpenAI raising API costs and with anthropic’s losses, how do they even keep going?
No way I can pay $500/month, or pay API equivalent pricing
2
4
292
I kinda wish this had been more clearly explained, but either way, Tibo takes a ton of heat from the users
Whether OpenAI delivers well or not, hope Tibo and all the team knows they’re appreciated
These guys and gals put a lot into pleasing us plebs and we’re insufferably demanding 😭😂
Hi,
Tomorrow we are re-opening the Pro $200 subscriptions to new subscribers, but together with it we are also changing how we calculate the usage for it. In effect, if you do the math, it will net out at half the dollar in API spend compared to the old Pro $200 plan.
Now that it's said, let me explain why this is happening and why you will still get more work done than if you were on the Pro $200 subscription one month ago.
(a) We didn't want to compromise in other ways and are committing to not reintroducing the 5h limit, so that you can fully use the weekly usage when you want.
(b) On the subscription, we guarantee that over time you always get more work done and with an increasing level of quality. This means that you will continue to get more value per dollar spent as a result of models getting more efficient and us passing down the improvements in the form of API price reductions.
(c) We don't want to put an incentive on ourselves to artificially inflate the API list prices to make it look like you are getting a lot (and workaround it through discounts, etc). Instead we want to continue to both rapidly reduce prices and increase capabilities of models on the API. This week we introduced GPT-6 Sol and GPT-6 Luna at 50% of their previous price. Over time, we see prices go low enough that it makes sense for most to buy usage as needed without there being a significant gap between what you get in a subscription and what you get in the API for a dollar spent.
(d) Tomorrow, we are adding more things to the subscription that won't draw on the usage, I won't reveal what that is yet.
I wanted to be transparent before all the big announcements tomorrow. Lots of new exciting things are coming to the subscriptions that will make it super compelling, but I wanted to make sure to share this change ahead of time so you can all understand it before we shower you with good news.
Codexingly,
Tibo
3
5
560
Random things are showing up on my feed and I’m not complaining
Life is so incredible
I just love it all
Serval is an African wild cat, fast and dangerous than snake
Community note
Servals are fast African wild cats that occasionally hunt snakes but are generally shy and not more dangerous than venomous snakes. The video shows a North American rattlesnake absent from the serval's African range.
nationalgeographic.com/animals/mammal…
ifaw.org/animals/serval…
en.wikipedia.org/wiki/Serval
animaldiversity.org/accounts/Lepta…
biologyinsights.com/where-do-rattl…
2
1
9
1,488
What are the odds models are trained to overdeliver when they’re first released, but only later on we see how they actually are?
Claude Sonnet 5.5 scored 70.6% on Terminal-Bench 4.0.
Then it got this prompt:
“Write HTML code that displays a 2D animation of a pelican riding a bicycle using SVG graphics. No additional testing is required.”
At xhigh: a solid SVG in 41 seconds for $0.057.
At max: 128K output tokens spent entirely on thinking, $1.28 gone, and no SVG.
Sometimes more reasoning just means failing more expensively.
3
2
131
AbsoluteWifey1776✝️🗽🇺🇸
I’m torn between being optimistic and pessimistic about DevDay
Part of me really wants them to make a comeback, and I think they can, esp with the “new thing” they’re teasing
HOWEVER, how has it not been leaked? If it was something so amazing, seems it’d have gotten out by now
2
6
398
Ok this is random… what if OpenAI does launch hardware??
33% chance OpenAI launches a consumer hardware product this year.
poly.market/Nue4P8W
1
1
160