BC

1K posts

BC banner
BC

BC

@buildwithbc

Quit to build my own startup | Building in AI

Beigetreten Şubat 2021
1.5K Folgt151 Follower
BC
BC@buildwithbc·
@alfongj x.com/ndrewpignanell… Got this as an answer from someone when I asked today
andrew pignanelli@ndrewpignanelli

Working on a similar problem in a sense. This type of thing is fairly complex behind the scenes. It spins up accounts at the partner providers, and some trusted platform has to handle payments/terms&conditions/data handling/permissions, which open source isn't really able to do at this point. Here, for example, supabase has to trust stripe to a level that they know stripe will ensure users signing up for provisioning also agree with supabase's terms of service, payment terms, and data handling.

English
0
0
0
137
Alfonso
Alfonso@alfongj·
why is this a stripe thing? is anyone building an open standard for this
Patrick Collison@patrickc

When @karpathy built MenuGen (karpathy.bearblog.dev/vibe-coding-me…), he said: "Vibe coding menugen was exhilarating and fun escapade as a local demo, but a bit of a painful slog as a deployed, real app. Building a modern app is a bit like assembling IKEA future. There are all these services, docs, API keys, configurations, dev/prod deployments, team and security features, rate limits, pricing tiers." We've all run into this issue when building with agents: you have to scurry off to establish accounts, clicking things in the browser as though it's the antediluvian days of 2023, in order to unblock its superintelligent progress. So we decided to build Stripe Projects to help agents instantly provision services from the CLI. For example, simply run: $ stripe projects add posthog/analytics And it'll create a PostHog account, get an API key, and (as needed) set up billing. Projects is launching today as a developer preview. You can register for access (we'll make it available to everyone soon) at projects.dev. We're also rolling out support for many new providers over the coming weeks. (Get in touch if you'd like to make your service available.) projects.dev

English
13
1
51
13.2K
Salina Mendoza
Salina Mendoza@inababi·
Is this a real email? Why would Claude need new permissions? @github Can we get better messaging please? @AnthropicAI @claudeai this is sketchy. Don’t make us look it up.
Salina Mendoza tweet media
English
1
0
0
89
BC
BC@buildwithbc·
@ndrewpignanelli @collision Yeah, that makes sense. I guess the angle that I was thinking was that it'd be better to be part of a wider protocol on how agents can spin up new services automatically
English
0
0
0
39
andrew pignanelli
andrew pignanelli@ndrewpignanelli·
Working on a similar problem in a sense. This type of thing is fairly complex behind the scenes. It spins up accounts at the partner providers, and some trusted platform has to handle payments/terms&conditions/data handling/permissions, which open source isn't really able to do at this point. Here, for example, supabase has to trust stripe to a level that they know stripe will ensure users signing up for provisioning also agree with supabase's terms of service, payment terms, and data handling.
English
1
0
2
181
John Collison
John Collison@collision·
Coding AIs can now provision real-world services with Stripe Projects: projects.dev
Patrick Collison@patrickc

When @karpathy built MenuGen (karpathy.bearblog.dev/vibe-coding-me…), he said: "Vibe coding menugen was exhilarating and fun escapade as a local demo, but a bit of a painful slog as a deployed, real app. Building a modern app is a bit like assembling IKEA future. There are all these services, docs, API keys, configurations, dev/prod deployments, team and security features, rate limits, pricing tiers." We've all run into this issue when building with agents: you have to scurry off to establish accounts, clicking things in the browser as though it's the antediluvian days of 2023, in order to unblock its superintelligent progress. So we decided to build Stripe Projects to help agents instantly provision services from the CLI. For example, simply run: $ stripe projects add posthog/analytics And it'll create a PostHog account, get an API key, and (as needed) set up billing. Projects is launching today as a developer preview. You can register for access (we'll make it available to everyone soon) at projects.dev. We're also rolling out support for many new providers over the coming weeks. (Get in touch if you'd like to make your service available.) projects.dev

English
14
14
249
74.9K
BC
BC@buildwithbc·
@levelsio Couldn't
English
0
0
0
3
Henrique Dubugras
Henrique Dubugras@hdubugras·
Now as an AoT-style anime, one of the best conversations from @hdinhdpod! How André Street built @SejaStone and reinvented Brazil’s financial system:
English
7
5
42
4.2K
BC
BC@buildwithbc·
@benln Where's Rio?
English
2
0
1
2.7K
Ben Lang
Ben Lang@benln·
Cursor team will be visiting these cities over next few weeks: • Bangalore • Chennai • Mumbai • Medellin • Cali • Bangkok • Istanbul • Singapore • Ho Chi Minh City • Jakarta • Kuala Lumpur • Colombo • Rio • Florianópolis Details being added on Luma!
English
431
146
4.4K
250.6K
BC
BC@buildwithbc·
@paulg @metapreston Hadn't seen this one before. Good read
English
0
0
1
1.1K
BC
BC@buildwithbc·
@paulg @TurnerNovak I actually thought they were twins for some reason
English
0
0
1
1.5K
Paul Graham
Paul Graham@paulg·
If you have multiple kids you might be surprised by how different they look.
Paul Graham tweet media
English
131
15
2.1K
297.1K
Minh-Phuc Tran
Minh-Phuc Tran@phuctm97·
AI is now really good at shipping MVP (solve real user problems, fully functional, can be quite complicated as well). Where AI still struggling big time is iterating & fixing bugs. AI doesn’t have the same memory as a human engineer / team does. It forgets its own code and needs to start from scratch again & again. This works at small scale but as the codebase grows, it becomes harder & harder for the AI to fully understand & keep track of what to account when adding new features / making changes. It eventually becomes impossible for the AI to avoid / fix its own bugs. And since everything was written entirely by AI from the beginning, we are no better than the AI at understanding that codebase. It becomes a dead end. This means, to truly embrace vibe coding, you need a solution for the AI to either remember or figure out how its new changes may impact the existing functionality and how to fix them without breaking other parts. Maybe AI memory becomes so good that it solves this on its own. Maybe it’s better dev tools for AI. For now, I find unit tests + integration tests are the best way to let AI test & fix its own changes.
English
18
2
30
4K
Ryan Carson
Ryan Carson@ryancarson·
The company that actually builds the agent-first code factory is going to be worth hundreds of billions of dollars. No one has cracked it yet. It can't be the model labs because then you're tied to one model. I'm hoping a company like @linear will do this. I'd happily pay thousands of dollars a month for that (+ the token cost). Basically, we need SDLC 2.0 for the agent age. (Also, the right solution can't rely on gh - we need whoever does this to completely replace it as well.)
English
122
14
349
55.7K
BC
BC@buildwithbc·
@OfficialLoganK Hey, one feedback for Gemini on the phone is that when you're doing a voice recording, any mini breath will auto send the message whilst chatgpt waits properly here. Not sure this is easy to adjust @OfficialLoganK
English
0
0
0
55
Logan Kilpatrick
Logan Kilpatrick@OfficialLoganK·
We just shipped a bunch of stuff to make it easier to scale with the Gemini API: - Automatic tier upgrades - Tier 1 -> Tier 2 now happens much faster (30 days post payment -> 3 days) and with less spend ($250 -> $100) - New billing account caps on each tier to limit over spend
English
95
35
907
79K
Tanner Linsley
Tanner Linsley@tannerlinsley·
Looking for the most reliable way to run agents in some kind of continuous/ralph loop mode, preferably overnight (@jamonholmgren). I'm use both Codex and Claude and am a sucker for OpenCode. What do I use?
English
69
4
228
53.9K
BC
BC@buildwithbc·
Tried implementing the concept of "audit (ralph) loop" to see how many times Claude will find something. Run "operator audit full" github.com/BVCampos/opera…
English
0
0
0
14
BC
BC@buildwithbc·
Given software code is fungible, opensourcing it ends up being the default
English
0
0
0
15
BC
BC@buildwithbc·
@andrewchen I was almost falling for this and getting one but I realized I probably carry my laptop to other places my foot pedal wouldn't go
English
0
0
0
43
andrew chen
andrew chen@andrewchen·
so i bought a USB foot pedal that triggers voice dictation for coding/email/whatever AMA
English
126
4
429
77.2K
BC
BC@buildwithbc·
@chongdashu Guess it's coffee time
English
0
0
0
47
Chong-U
Chong-U@chongdashu·
■ {"detail":"Bad Request"}
English
3
1
10
1.9K
abdumajid
abdumajid@abdumajid_3001·
I can't login to claude code from an hour, is it working? @trq212
abdumajid tweet media
English
15
0
156
18.9K

Entdecken

@alfongj @inababi @github @AnthropicAI @claudeai @ndrewpignanelli @collision @levelsio