Twitter Trends April 28, 2026

Claude now connects to the tools creative professionals already use. With the new Blender connector, you can debug a scene, build new tools...

Earlier this month, an Erdős problem that had been open for 60 years was solved with help from GPT-5.4 Pro. What happens now that AI is get...

There's a quadrillion-dollar question at the heart of AI: Why are humans so much more sample efficient compared to LLM? There are three poss...

Y Combinator's Summer 2026 request for startups:

You're going to be embarrassed by the slides you made before AI Meet Replit Slides The first AI slides with stunning design

deepseek is outrageously cheap even if you ignore the current 75% discount input token prices are 35x cheaper than opus cached tokens are ...

"I keep meaning to really dig into AI, but I can't seem to find the time. I'm just so busy." If that sounds like you, @yourgirlhils and I m...

I have Claude Code fully integrated into my smart glasses Everywhere I go I am now vibe coding When you are talking to me in real life, I ...

This week we opened Lovable's new Stockholm HQ. Since the beginning, we've built Lovable on the thesis that in-person is a competitive adva...

gogcli 0.14.0 🧭 🔐 Encrypted Google backups (local+encrypted Git) 📬 Resumable Gmail + Drive export 📝 Markdown Gmail mirrors 📇 vCard con...

In order to use LM Studio with my local models on my local network, I have to apply to use "LM Link" and hope they get me off the waitlist? ...

im writing an essay in proof using codex's in-app browser. i type directly in the document, and codex loops in parallel on the left, collab...

Real quick, here's how you can turn a design element into a skill and reuse that. DESIGN.md, Skills and HTMLs are incredible prompt referen...

Our approach to hiring is a little different. We value slope over credentials, judgment over polish, and real work over interviews. Here’s...

Huge news: I'm starting a /skills site on AI Hero about my skills repo - still the top of GitHub trending 🤯 Sign up for skill updates, vi...

one of the coolest things (also detrimental in some ways) about the internet is the optionality it gives you. to become extraordinary, you ...

The last month, Anthropic: - Quietly nerfed their flagship model harness (Claude Code) without telling anyone - Banned corporate customers...

A Google Earth + Flight Simulator, fully cooked by Perplexity Computer with Codex/CC as subagents inside it. GTA vibes.

every prediction i see on here about what the future of software looks like is just programmers finding an excuse to not have to do product ...

President of @Blackstone Jon Gray says LLM spend among the company's portcos is up 15-fold in Q1 of this year over last year. At the same t...

Here's a fun skill you can teach your OpenClaw/Hermes, just paste it and find a book and it'll give you hours of fun Book Mirror — How It W...

“YouTube videos about AI, but with no soyface”

We just tagged a new release to make it easier to work with presets. We've also added monorepo support for shadcn apply.

Thinking about recording some videos live on stream tomorrow, @theo-style What topics should I cover? What videos do you want to see?

I want to rename /domain-model before I make a video about it. It's a drop-in replacement for /grill-me that's better when you're working w...

BREAKING: Eric Romo (@_Eric_Romo), President & COO of Impulse Space (@GoToImpulse) says: "The idea that launch prices are going to fall si...

Jim Simons returned 66% per year for 30 years. Net of fees, 39%. The Medallion Fund is the best track record in financial history, and it is...

The FDE is the most essential role in tech and AI adoption today. There is no market for generalist enterprise products anymore. If you han...

using posthog inside of codex it is SO MUCH better to browse posthog inside of codex—codex can write queries and see the results, and it ma...

since openai is building a phone, let’s think a little bit about one of the core aspects of a phone which is the messaging experience. the ...

How to improve any software system: 1. Do less work 2. Do the work faster 3. Do the work in parallel (or elsewhere)

Don't get AI to generate images; get them to generate SVGs! Vector illustrations that seamlessly blend into the design style is the perfect...

President of @blackstone Jon Gray says his firm's biggest theme after data centers is electricity: "Data centers, robots, autonomous vehicl...

First company to solve the AI cost problem will be the next tech giant for the next two decades at least.

Please, pretty please, *actually* test restoring from your db backup (which, for the love of god, should be from an offsite location).

Shipped an update to my skills that allows you to use them with any issue tracker (GitHub, Jira, Linear, local files) 1. Run npx skills@lat...

this is average spend per session with each new model release people are spending more and more

A big problem with all AI at work punditry right now is that it all rests on data from the pre-agentic era (which is basically just now endi...

The new LLM trained only on pre-1931 text is small enough that it can potentially run on device, so, with the right tools, you can get a ful...

BREAKING: Rainmaker plans on tackling desalination. Their 3-year roadmap, per @ADoricko: 2026: Master snow precipitation enhancement 2027...

MCP updates in FigJam so you can visualize your systems (and not just read code) → generate_diagram to create architecture diagrams & ERDs ...

THE 64 MINUTE OPENAI CODEX MASTERCLASS IS HERE if you've been meaning to learn Codex, this is the episode for you, we cover: 1/ Claude vs ...

.@bubbleboi says we're still at the very beginning of the AI cycle and have at least three years of "capex going crazy": "At the beginning ...

how do I disable auto-scroll in Codex and Cursor? I thought we had this.

our goal with @skye is to build deeply *emotional* products that also provide real utility. this distinction matters a fuck ton because so ...

Little thought experiment to put AI chip improvements in perspective: Imagine that every person on Earth uses a calculator to perform one ...

Are you missing out on the next 10x engineer? Saffron (@trysaffron) lets companies evaluate how well software engineers code with AI tools. ...

Supabase is now an official ChatGPT app!

.@michael_nielsen made an incredibly interesting argument about why the tech tree is actually way larger than we realize, and how our descen...

Codex is doing AI research for me on Talkie using its in-app browser research question: can an LLM trained only on 1930s text learn to code...

This chart - coming from GitHub is hard to interpret, thanks to missing information. 1. Where does this chart start? Aka what are the valu...

Will experiment with https://github.com/brunoborges/ghx since I constantly run into GitHub rate limit issues. Team there does the best to he...

BREAKING NEWS: BOS and NY Tech Week calendars are LIVE! You can now register for events: [http://tech-week.com/calendar] 2,000+ event subm...

Hey @Calendly - It's super annoying that my claw can't book a slot on someone's Calendly. Wake up please. We're in an agent-first world n...

I've had a secret agenda with my podcast: learn what I want to. 145 episodes in, I had a bunch of fun looking at the neural network I'm for...

Is anyone building something to make it easier to manage write-offs for angel investments?

A venture firm saying we should have a podcast is like when they say they think AI is going to be big.

getting to inbox 0 with Codex + @CoraComputer inbox codex running in a loop on the left, Cora open in a codex's browser on the right this ...

.@bubbleboi's new thesis for Intel is all about advanced packaging, which he says is the new Moore's Law: "Up until this point, Intel's the...

I have been playing with the new Outlook agent, and it is fine, but really awkward to use, since you have to ask for things in a chatbot win...

Some notes on talkie, a new "vintage language model" from a team including Alec Radford (yes, that Alec Radford) "trained on 260B tokens of ...

Introducing OSSCAR, with @commitvc. An index of the fastest-growing open-source orgs. #1 Emerging: @papercliping (autonomous business buil...

Two Korean companies are now worth more than every technology stock in the Stoxx Europe 600 combined. That's the reason South Korea's stock ...

People who are sharing their take about reliability at GH: - The COO (owning operations) - The CTO (engineering) - The CPO (product) The...

21 days until Google I/O ⏳ Three weeks. That's it. 🔵🔴🟡🟢

.@Blackstone President Jon Gray says that just like Sears and other retailers went out of business in the dot-com era, but companies like Co...

I'm again rate limited on GitHub, but codex just opened the browser and clicks around GitHub as workaround.

parallel video processing in the browser audio gets ripped & compressed via Mediabunny, then sent to @AssemblyAI for processing result is ...

.@Blackstone President Jon Gray explains his approach to theme investing while staying away from high multiples: "The way to think about it...

A meta-analysis of 3,075 college students just settled the laptop vs paper debate. Handwriting won. The effect size: r = -0.142 favoring pe...

Linear Agent x @glean Pull in company knowledge and context from Glean to help Linear Agent generate sharper project plans and updates.

.@ADoricko says that Rainmaker has already made cloud seeding precise down to ~8,000 square acres and believes it can get that down to ~1,00...

Every “working” company in enterprise AI is doing the same thing - completing valuable tasks in a verifiable way - Valuable tasks = cost sa...

Replit + Security | Community AMA with CTO Luis Héctor Chávez https://x.com/i/broadcasts/1YxNrZYVeoZxw

I give it 2 weeks before someone in SF starts selling peptides for agents

you don’t realize how CHEAP DeepSeek is until you use it all day and pay the price of a bag of chips. DeepSeek V4 Pro: ~$0.87 / 1M tokens O...

This year's AI Accelerator winners shipped agents for finance teams, application security, and real estate. They built with $8M in credits ...

"I fucking HATE AI" the harsh reality from a developer with 20 years of experience, on the existential dread to see decades of skills slowl...

OpenAI's CFO just told her own executives the company might not be able to pay its $1.4 trillion in compute bills. OpenAI makes $20 billion ...

Learn how to run a local coding agent! Use: - Pi agent - Gemma 4 26B - Serving engine of choice: e.g. LM Studio

ARC-AGI-3 testing is done for gpt-5.5 and opus 4.7 Now we’re in analysis mode going through the logs It’s pretty clear where the failure m...

The past week I've recorded so many pod eps covereing a wild range of usecases - codex + tmux for DB work - the vibiest of vibecoders making...

new podcast pages on @beehiiv... > distro with Apple, Spotify, Overcast, etc. > dedicated page for each new episode > auto-generated transc...

What if we measure how good AI models are by what era of games it can one shot a single level for (or maybe build with 30 min of work): Ata...

Has anyone gotten clean, seamless clip-to-clip extends working in Seedance 2.0? How? By seamless, I mean: - no camera cuts - no visual or a...

From the stage at AI Dev 26: Dan Maloney, CEO of @LandingAI, on the Future of Software Engineering: AI is compressing skill gaps. Whether y...

What if I want my coding agent to mention goblins? (If you don't know the context for this, I suspect it will become viral soon enough)

I hacked on some hardware to solve my late-night ideas problem: You ever have late-night ideas, but don't want to use your phone??? It'll ...

The most sophisticated AI users I know are talking to just ONE agent daily, not multiple agents That agents will then assign tasks to other...

AI wrapper - lame, undifferentiated, commoditized AI harness - mysterious, technical, strategic

I am worried by the constraining exit options for venture backed companies: 1. IPOs: You can’t IPO with less than $500M in revenue and eve...

Emma McGrattan, CTO of @ActianCorp, explored at AI Dev 26 how the shift toward distributed AI is changing the role of vector databases. Her ...

true life: I'm addicted to using the @beehiiv MCP to extract insights and build dashboards

ARC Prize 2026: ARC-AGI-3 now has H100s Thank you to our partners at @kaggle for powering the ARC-AGI-3 competition with H100s Participant...

Eli Schilling's workshop at AI Dev 26 focused on Memory Engineering and Context Engineering, and shared a working mental model of both disci...

what would a "better github" even look like?

An easy way to get a team engaged with AI is just to build the thing you are talking about in the meeting during the meeting using Codex or ...

Nvidia is now bigger than Japan. In the MSCI All Country World Index, the global benchmark that ~85% of cross-border equity money tracks, o...

"If an automation doesn't work 100% of the time, it's not really an automation." -- Claude Code and Cowork Head of Product Cat Wu cc: @Anth...

Ineffable Intelligence, China Blocks Manus, OpenAI Loosen Microsoft, Palantir Joins https://x.com/i/broadcasts/1AGRnadVeEaGl