LightNews LightNews
banner
twittertechposts.bsky.social
@twittertechposts.bsky.social
1 followers 8 following 49 posts
This is a curated list of X posts from tech-related people who are not posting on BlueSky. Let us know if you miss someone from there, so we can add them here. Let us know if any of these people starts activelly using bsky, so we stop sharing their posts
Posts Replies Media Videos
Pinned
twittertechposts.bsky.social
twittertechposts.bsky.social @twittertechposts.bsky.social · 20d
This account was created to share a curated list of X posts from tech-related people who are not activelly posting on BlueSky. If you miss someone from there, write it down here so we can add them. If any of these people starts activelly using bsky, write it down here as well.
twittertechposts.bsky.social
#ai #Microsoft
x.com/dwarkesh_sp/...
@dwarkesh.bsky.social
.@satyanadella gave me and @dylan522p an exclusive tour of Fairwater 2, the most powerful AI datacenter in the world. We then chatted through Satya's vision for Microsoft in a world with AGI. 0:00:00 - Fairwater 2 0:04:15 - Business models for AGI 0:13:42 - Copilot 0:20:56 - Whose margins will expand most? 0:37:12 - MAI 0:48:42 - The hyperscale business 1:03:39 - In-house chip & OpenAI partnership 1:10:30 - The CAPEX explosion 1:16:01 - Will the world trust US companies to lead AI? Look up Dwarkesh Podcast on Youtube, Apple Podcasts or Spotify to tune in
November 12, 2025 at 7:46 PM
twittertechposts.bsky.social
#shadcn
x.com/shadcn/statu...
www.youtube.com/watch?v=_kVx...
@shadcn.com @thekitze.bsky.social
This is one of my favorite talk. You should watch it. It covers everything from fomo, hype cycles, jumping on shiny new things, abstraction, copy/paste, AI and generated code. Seven years ago! It's by @thekitze . Still hits hard. Watch it.
November 12, 2025 at 7:08 PM
twittertechposts.bsky.social
#design
@leerob.com
x.com/leerob/statu...
Lee Robinson The most practical design advice I've heard is to constrain yourself. Pick one or two font sizes. Use only a few grays. Add color sparingly. Most of my old designs with three font weights, six grays, and four different text sizes all looked instantly better when stripped back.
November 12, 2025 at 7:06 PM
twittertechposts.bsky.social
@leerob.com
x.com/leerob/statu...
The launch video hype cycle is almost completely over 2026 is when the blog post returns. Words are so back
November 6, 2025 at 8:22 AM
twittertechposts.bsky.social
#cursor #ai #git #ai-agent
@leerob.com
x.com/leerob/statu...
Cursor now natively supports git worktrees. You can run multiple agents in parallel, each with their own isolated copy of the codebase. This is great for comparing models or running the same model multiple times and picking the best result.
November 6, 2025 at 8:20 AM
twittertechposts.bsky.social
#lexfridman
@lexfridman.bsky.social
x.com/lexfridman/s...
Lex Fridman @lexfridman · 9h Over the past several weeks, I've been back in the lab working crazy hours like old times, and loving life. This was much needed. I'm happy ;-) If I'm not programming, hands-on with robots/hardware for many hours a day, I feel a bit lost and less able to think clearly about the world. It's just how my brain is built. I think it's because engineering humbles me to the reality of the world. I need both literature & engineering in my life 🤣 I love both, and both are a source of truth. I'm still doing the podcast, just adding much more engineering back into the mix ;-) The podcast gives me an opportunity to celebrate great scientists, engineers, and builders. This is one of the things I love doing most. In the coming months, I'll continue splitting time between MIT and Caltech (Boston & LA) working with some brilliant people. The focus of the work is human-robot interaction with quadrupeds and humanoid robots. I'll also likely be at NeurIPS in San Diego this year. If you want to chat over coffee in Boston, LA, or San Diego for NeurIPS or have guest suggestions in those places, please fill out the coffee or guest forms respectively (see link in comment). For NeurIPS, please choose the "NeurIPS 2025" option for special event in the form.
November 6, 2025 at 8:12 AM
twittertechposts.bsky.social
#shadcn #vscode #cursor
@shadcn.com
x.com/shadcn/statu...
TIL: In VS Code/Cursor, use "Reset Editor Group Sizes" to make split panes equal width.
November 6, 2025 at 8:07 AM
twittertechposts.bsky.social
#coding
x.com/leerob/statu...
@leerob.com
leerob.com/personal-sof...
Personal software is great. You can build small helpful things. There are so many janky websites littered with ads and cookie banners. So I make my own utilities instead. This took two prompts. And I can design them exactly how I want and own the code forever.
November 3, 2025 at 8:25 AM
twittertechposts.bsky.social
#antropic #openai #llm #ai
x.com/rohanpaul_ai...
@rohanpaul.bsky.social
Anthropic has overtaken OpenAI in enterprise LLM API market share. OpenAI fell from 50% in late 2023 to 25% by mid-2025, which shows that brand alone does not hold share once real workloads start. Anthropic now leads enterprise LLM API usage with 32%, while OpenAI has 25%, pointing to a real shift in how companies pick vendors. Enterprise LLM API spend hit $8.4B in the first half of 2025. Anthropic’s push on data controls, compliance, and clean integration with existing systems won trust, and that trust tends to decide renewals and expansions. Claude’s recent lines, including stronger reasoning and coding, helped too, with developer code-gen share around 42% for Anthropic vs 21% for OpenAI. Usage is shifting to inference at scale, so uptime, latency, and incident response matter more than raw benchmark wins. Vendor switching stayed low at 11%, and 66% of teams just upgraded within the same vendor, so any share gain here is hard won. Google sits near 20% and Meta near 9%, so this is not a 2-player market, and strengths differ by use case like agents, code, or retrieval. Buyers now weigh cost per token, data residency, auditability, SOC reports, and fine-grained controls as much as model quality. Multi-vendor setups are rising because they reduce lock-in and let teams route tasks to the best model for that job.
November 3, 2025 at 8:18 AM
twittertechposts.bsky.social
#shadcn #tanstack
@shadcn.com
x.com/shadcn/statu...
Tanstack Start + shadcn/ui is now just one command.
November 3, 2025 at 8:16 AM
twittertechposts.bsky.social
#vibe-coding
@officiallogank.bsky.social
x.com/OfficialLoga...
Logan Kilpatrick @OfficialLoganK vibe coding should be taught in every school and university around the world, first day on campus, welcome to the future. Logan Kilpatrick @OfficialLoganK my predictions if this happened: - students building companies goes way up - student debt over time goes down - CS enrollments flat or up - dropout rate nominally increases - student engagement with class material goes way up (interactive courses) - STEM dropout rate down
November 3, 2025 at 8:13 AM
twittertechposts.bsky.social
#shadcn
@shadcn.com
x.com/shadcn/statu...
Today's @uidotdev newsletter is about shadcn/ui ;)
November 3, 2025 at 8:11 AM
twittertechposts.bsky.social
@lexfridman.bsky.social
x.com/lexfridman/s...
November 3, 2025 at 8:09 AM
twittertechposts.bsky.social
#cursor #ai #coding
@leerob.com
x.com/leerob/statu...
Lee Robinson @leerob · Oct 31 Cursor 2.0 can now: 1. Use any frontier model or our own agent model 2. Run multi-agents with built-in git worktrees 3. Create plans with one model, implement with another 4. Review + create a PR in one click 5. Close your laptop and handoff to the cloud
November 3, 2025 at 8:07 AM
twittertechposts.bsky.social
#ai #artificialanalysis #llm #benchmark
x.com/ArtificialAn...
Big technology companies are continuing to play across all AI modalities while smaller challengers tend to focus on specific modalities. In language models, OpenAI has reclaimed the top spot for the most intelligent model with the release of GPT-5 (high), however with frequent model releases by all frontier labs, the race is closer than ever. The United States 🇺🇸 and China 🇨🇳 remain dominant in AI with few models coming from other companies. See more in our latest Quarterly State of AI report on our website (link below)
November 3, 2025 at 8:04 AM
twittertechposts.bsky.social
#cursor #ai #coding
@leerob.com
x.com/leerob/statu...
Lee Robinson Cursor now has a built-in web browser. You can say "start my app" and it will run the dev server in a terminal and then navigate the browser to localhost:3000. It can take screenshots of the page for iterating on design, and you can send DOM elements to the agent with prompts!
November 3, 2025 at 8:01 AM
twittertechposts.bsky.social
#shadcn #mcp
@shadcn.com
x.com/shadcn/statu...
see video in the comments
shadcn You can now configure MCP right from the directory page. Copy to your components.json file.
November 3, 2025 at 7:56 AM
twittertechposts.bsky.social
#typescript #js #ts #programming
@rauchg.blue
x.com/rauchg/statu...
The world runs on TypeScript & JavaScript. Our bet is that AI engineering will follow suit. The growth in @aisdk downloads and adoption has been astonishing. When we wrote the Ship AI keynote it was at 3.4M weekly downloads. A couple weeks later, it’s now at 4.1M
October 30, 2025 at 11:23 PM
twittertechposts.bsky.social
#cursor #ai
@leerob.com
x.com/leerob/statu...
October 30, 2025 at 11:21 PM
twittertechposts.bsky.social
#ai #llm #api #googleaistudio
@officiallogank.bsky.social
x.com/OfficialLoga...
October 30, 2025 at 11:19 PM
twittertechposts.bsky.social
#ai #llm #artificialanalysis #text2video #veo3.1
x.com/ArtificialAn...

See Veo 3.1 for yourself in the Artificial Analysis Video Arena: artificialanalysis.ai/text-to-vide...
October 30, 2025 at 11:16 PM
twittertechposts.bsky.social
#ai #llm #artificialanalysis #text2speech #sonic3
x.com/ArtificialAn...

Explore different models and voices in the Artificial Analysis Speech Generation Explorer: artificialanalysis.ai/text-to-spee...

Vote for models in the Artificial Analysis Speech Arena:
Cartesia has just released Sonic-3, their new flagship Text to Speech model for realtime conversation Sonic-3 delivers Text to Speech streaming with built-in prosodic elements such as laughter, as well as emotions like surprise and curiosity. The new model supports 42 languages, including 9 Indian languages. Sonic-3 leverages a State Space Model (SSM) architecture like Cartesia’s previous releases. Cartesia has been a leader in using SSMs, with other models typically using transformer-based architectures. SSMs are generally more compute and memory efficient, which likely contributes to Sonic-3’s comparatively impressive latency. Cartesia offers a subscription and credits pricing model ranging from $0 per month for personal use, up to $299 per month for their Scale package. Sonic-3 has been added to the Artificial Analysis Text to Speech Arena for blind preference voting.
October 30, 2025 at 11:14 PM
twittertechposts.bsky.social
#ai #llm #benchmarks #granite-4.0-nano
x.com/ArtificialAn...
IBM just released Granite 4.0 Nano, a family of four tiny open weights models (1B and 350M) focused on efficiency with strong performance for their size Granite 4.0 H 1B and Granite 4.0 1B score 14 and 13 on the Artificial Analysis Intelligence Index, while the 350M variants…
October 30, 2025 at 11:07 PM
twittertechposts.bsky.social
#shadcn #mcp
@shadcn.com
x.com/shadcn/statu...
shadcn Yes this works. You can ask the shadcn mcp to mix and match code from registries. Here I asked for the confirmation component from ai-elements (radix), but told it to use primitives from coss-ui (base). Use the shadcn mcp to add the confirmation component from @ai-elements but use the primitives from @coss. Searching for a confirmation component in @ai- elements using the shadcn MCP, then adding it using primitives from @coss.
October 29, 2025 at 10:00 PM
twittertechposts.bsky.social
#vercel #bun
@vercel.com
x.com/vercel/statu...
vercel.com/blog/bun-run...
Vercel We now support Bun on Vercel. Bun is an open runtime pushing the frontier on DX and performance. Start building and see our latest benchmarks.
October 29, 2025 at 7:28 AM