पिन किया गया ट्वीट
john wayne
603 posts
@jhleath @LuchoMzmz Is file system not the best storage system? Idk, maybe expose confluence as a file system to the agent?
English
it's actually quite simple, you need to own the storage layer.
if you own the part of the software stack that owns the answer (from the Kernel) to the question of "what exists inside this directory" and "what exists inside this file", then it's reasonable to be able to materialize data into the agent using any number of mechanisms
i believe that only the authors of the agent know the best way to represent data to their agent, so one-size fits all (like MCP from the system of record itself) isn't going to cut it for the agents being built on the frontier
English
it remains super odd to me that none of the existing agent frameworks, Mastra, Flue, or now Eve seem to do anything about getting context into the agent?
every team that i talk to who are designing agents at-scale need to figure out how to get the enterprise data *to* the agent, which requires carefully planning ETL, evaluating how well the agent performs with different data formats, running things like map-reduce
and yet, every one of the agent frameworks just... leaves this to user with no opinion on it?
maybe you can only do this part with a fully-owned storage layer, idk
Vercel@vercel
Introducing eve, an agent framework. 𝚊𝚐𝚎𝚗𝚝/ 𝚊𝚐𝚎𝚗𝚝.𝚝𝚜 𝚒𝚗𝚜𝚝𝚛𝚞𝚌𝚝𝚒𝚘𝚗𝚜.𝚖𝚍 𝚝𝚘𝚘𝚕𝚜/ 𝚜𝚔𝚒𝚕𝚕𝚜/ 𝚜𝚊𝚗𝚍𝚋𝚘𝚡/ 𝚜𝚌𝚑𝚎𝚍𝚞𝚕𝚎𝚜/ Like Next.js, for agents. vercel.com/blog/introduci…
English
I feel I’d be more interested in work if my company’s codebase wasn’t garbage
English
@infinitefun_ @pmarca Never ending business model, they get to sell us next years slop each year
English
most of what is considered "taste" (read: design) is in the realm of zero sum signaling games
your tasteslop will just be the next AI slop
and then your anti-tasteslop will become the next tasteslop
taste is defined in terms of slop and therefore can never transcend slop.
Thais Castello Branco@thaiscbranco_
We’re excited to introduce Taste Labs. Our mission is to end AI slop. We’re building the data and infrastructure layer to give AI models and agents taste. And today we’re coming out of stealth, announcing our $18.5M seed funding, co-led by @CRV and @AmplifyPartners AI has nailed objective domains and made it easy to generate anything. But it still feels off. Now, the challenge is judgement. What fits, what feels like you, what’s GREAT. This requires turning a fuzzy, subjective domain into something we can measure and codify. We’re starting with design. There are two sides to cracking this, the foundation model layer and the agent layer: - We’ve already been working with the top frontier labs to evaluate and improve their models, crafting the right post-training data and RL environments. - We’ve also been working with app-layer companies to build the context and verification tools for their agents to produce better, more on-brand, more creative outputs. We want a future where AI feels right. If you’re passionate about this mission, join us!
English
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: anthropic.com/news/fable-myt…
English
@thsottiaux @robertcourson Is there learning curve to optimizing a new pretrained model like spud compared to the previous base model?
English
at this point I don't want GPT 5.6 - it needs to be GPT 6.
No way Anthropic has completely blown past them like this.. 3 models in 2 months and Fable is not even their best model. Feels like anthropic ruined OpenAI's whole model roadmap/release plan.
English
English
I THOUGHT THE IPHONE 16 WAS BUILT FROM THE GROUND UP FOR AI?
Apple not supporting everything on their less than 2 year old AI devices is insane.
MacRumorsLive@macrumorslive
Apple Intelligence features are available on iPhone 15 Pro and later, but the most powerful on-device model is limited to newer iPhones, iPads, and Macs. Developers are getting access to Siri AI today, but it is not going to be available in the EU initially.
English
@ericzakariasson Been using cursor at work for some iOS development and nearly all models quit mid run using Cursor 3. This never happened to me before on personal projects but it happens every session at work.
I’m using typical settings, tried Composer 2.5 and GPT 5.4/5.5. Ideas?
English
lmao all of our designers got one of their apple displays replaced by the cheapest one from amazon
in pursuit of higher contrast
English
Highest non-OpenAI-employee Codex usage I've seen on the timeline so far 👀
On-par with the Codexmaxxer @jxnlco himself
@OpenAIDevs keep cooking 👨🍳
English
CODEX LIMITS REDUCED 50%
can someone verify if this is true ? @thsottiaux
English
@steipete I've been using clawpatch inside my /goal loop and it's been amazing.
Do you have anything I can read to try to reproduce your dev style? I got 200x codex plan rn (thx 5.5) and want to push my tokens to the limit
English
@Michaelzsguo I used some other OS models in the codex harness and they couldn't run properly in the harness. Terminated mid sentence, unable to call tools, etc.
You sure it's running properly?
English
My Deepseek V4 Pro agent (inside codex) has been pursuing goal for more than 13 hours, burning ~100M tokens, and has only costed me $1.85. Yes you saw it right. Not $185, but 1 dollar 85 cents.
English
@thdxr happened to me yesterday, switching back to 5.4 helped
English
In the limit, everything becomes a gambling app. So sad
Tuyo@itstuyo
We created a card that sometimes doesn't charge you. Buy Now, Pay Maybe.
English
@SinfulMeatStick @techbrewmb Tbf you still have a working phone, Siri just sucks
English
@techbrewmb Pay a $250 million in fines while making billions in the actual fraud.
That’s crazy work
English
Apple will pay $250 million to settle a class-action suit accusing it of misleading US customers about AI features on new iPhones
Who's eligible:
• Any iPhone 16 model
• iPhone 15 Pro, Pro Max
• Purchased between 6/10/24 - 3/29/25
That comes out to roughly 37 million devices
English
@conacts1 It’s called Agent Flywheel, try it out (it’s free and open source):
agent-flywheel.com
English
Damn, that 10x usage gift for the OpenAI party rejects would sure come in handy now. Guess I need to start adding more accounts 😭
English