Novalede: https://nitter.cf/t.co/pwFfWyVz15 Flowtelic: https://nitter.cf/t.co/wKHOOSb1VY YouTube: https://nitter.cf/t.co/LZMdwtWDxF
Mauritius
Joined June 2009
- Tweets42.3K
- Following2.3K
- Followers4.6K
- Likes71.6K
Pinned Tweet
Introducing Novalede: Your world, edited.
novalede.com
Get your inbox, events, favourite news, alerts, and business stats written like a newspaper with a Morning and Evening edition. Stop important emails being buried, and have the lede of the story right there front and center.
Add it to your favourite agent as an MCP connector and you decide what you want it to publish.
Waitlist only, free to use, with a few seats released at a time. Reserve your spot now.
Having AI build things in Blender is such the superpower I've wanted for years. I've always wanted to learn Blender and even did the donut course, but it just didn't stick because I don't do it often.
But with AI, I can now explore creating things with it and see what happens.
Grok Bot ➡ Nerfed usage
Claude Duet ➡ Waitlist
ChatGPT Dots ➡ Waitlist
Meta Muse ➡ Waitlist
So the cost of Grok Bot feels like it doubled. Or more to the point, for what I can see, they recalcualted people's usage at being 2x. People who were on 60% seemed to jump to 100% used. I was on 24% and it jumped to 49%.
What I don't get is:
a) Why they didn't warn us this was happening?
b) Why not do the switch at the point people's usage resset?
I do wonder now if the new plans already had this usage baked in, and it's only those on the legacy Cursor plans who got hit. Does it make switching to the new plan more like-for-like?
So the cost of Grok Bot feels like it doubled. Or more to the point, for what I can see, they recalcualted people's usage at being 2x. People who were on 60% seemed to jump to 100% used. I was on 24% and it jumped to 49%.
What I don't get is:
a) Why they didn't warn us this was happening?
b) Why not do the switch at the point people's usage resset?
I do wonder now if the new plans already had this usage baked in, and it's only those on the legacy Cursor plans who got hit. Does it make switching to the new plan more like-for-like?
The problem with agents and coding using the same shared pool is akin to not being able to email your staff because their computer has compiled too much code. It makes no sense.
Let’s hope that anyone who has seen their Grok Bot usage jump without using it much is going to be fixed by this…
Anyway else get the feeling that the AI subscription honeymoon is coming to an end.
Cursor ➡ API rates for non-Grok models
Claude ➡ Removal of the 20x plan
ChatGPT ➡ Codex and ChatGPT now have the same limit, so your coding agent can cut off your web searches
I woke up this morning with my Grok Bot alerting me about a Novalede story for the Claude extension backdoor in the Cursor marketplace, but goes one step further and tells me how to confirm if my Mac is infected.
I love the proactive nature of this agent!
Yes, I have a nice web app now that gives me push notifications when my Grok Bot wants to access my servers. So it means that is temporary and usage specific. It uses with Teleport (self hosted) with a custom security token page to unlock authorisation. Then Grok Bot gets short lived access to that it needs.
Okay, Microsoft cooked with this on Microsoft Clarity 👏👏👏
@msftClarity no cap, you ate and left no crumbs 🫡
I quite often find myself writing a thoughtful reply here on 𝕏 only to discard it thinking, ‘I don’t want the hassle’
I really don't understand Apple's new Macbook Pro lineup.
M6
M5 Pro
M5 Max
I don't know if the M6 is supposed to be faster than the M5 Pro or M5 Max - I guess the memory bandwidth is different but overall, is the M6 faster.
And their own benchmarks show the M5 Pro being faster at 3D rendering than the M5 Max.
I just wish they compared them against each other.
