@mattpocockuk I ended up renting a machine based on hearing @thdxr on @syntaxfm . It makes life so much easier without OOM or disk space issues then have @Tailscale serve a port range. Never been so much more productive.
The best thing about Computer in Codex is having it deal with certs and all the other stuff to get an App on TestFlight with iTunes Connect and the Google Play Store.
I found myself that I’m like mostly compute constrained and it seems like doing some more memory optimization will be helpful, especially if you’re like you running multiple concurrent agents at the same time. It’ll be great if you would like, somehow offload some of those things and handle and handle that directly.
The arrangement with the team was to take calls and standups super early, before most people would wake up and then use the 9am to 12pm block however folks wanted. This did included Saturday too.
It was not a boondoggle, those morning calls/meetings were intense. After being screamed at for 3 hours straight you needed that break to survive.
@AlexFinn 1000% such a great feature. being able to have an orchestrator construct a graph to work through tasks is a game changer.
also you should try using it with multiple people go for a walk and have two or more people riff on an idea togethe.
Today, we’re launching Bland Speech v3 - The world's first Human Speech Engine.
In @designarena’s Audio Realism benchmark, Speech v3 is the top model, outranking Elevenlabs, Grok, Cartesia, and OpenAI.
Trained on over 100 million real, human conversations: businesses, developers, and creators can now create any audio they can imagine.
To showcase this, we used 5 seconds of old footage to help James, a 49 year old father who recently suffered a stroke, get his voice back.
Try it for free at https://t.co/lr96zcKLbq
Full length documentary below🧵
@CloudflareDev Can we pls just get budget controls at the product/account level?
Otherwise have to use this API with computer use to go to chrome and make the changes.
Use codex and have an orchestrator manage all the threads/agents. Then have that thread manage all the other threads.
You can just have it manage everything and then check in like a micro manager ever couple of hours, “what are you doing” “what needs help” “what have you done today”….
This is a good ad, but I still think that AI is accelerating everything.
The best A/C companies are telling people to use ChatGPT, had an issue with a Mitsubishi Heat Pump and the guy was like, I can come out for $900 or you can just go take a photo of the serial number in your attic and ask ChatGPT how to do a soft reset.
Same with plumbing. I was quoted $1k to replace a garbage disposal. I used ChatGPT and spent $150 and did it myself.
I used to tell Codex what to do. Now you can literally tell Codex what to do.
You can use your voice to orchestrate tasks with larger models, check progress and steer the work all by talking on your laptop or on your phone. It's WILD!
@mattshumer_ Curious how much RAM you need to run this? I feel like my biggest battle with queries like this is that I’m always hardware constrained because the harnesses don’t seem to be good optimizing performance.