Selina Wang retweetledi
Selina Wang
285 posts
Why hasn’t @dylan522p started a hedge fund yet?
Patrick OShaughnessy@patrick_oshag
Dylan on why memory prices will double again: "Memory can only grow capacity low double digit percentages a year. 20-30% a year, even less for NAND, a little bit higher for DRAM, but whatever. Even though the demand signal was very strong at the end of 2025, the memory companies immediately started reacting. None of that incremental capacity really gets here until 2028, even if they wanted to build as fast as possible. So the result is memory prices have gone through the roof. They're going to double and triple again. DRAM, especially. People are like, "oh, the memory story is overplayed." No, you don't get it. DRAM will double or triple from here still, because that's how much capacity is required. And they have to steal capacity from somewhere else. And the only way to steal capacity from somewhere else in a capitalist economy is demand destruction via higher pricing."
English
Forward deployed engineers have become one of the most sought after hires in AI📈. The FDE boom is real, but also a sign of tech immaturity...
We've seen this before:
- Early internet needed web consultants for things that became drag-and-drop
- Early cloud needed solution architects for things that became one-click
- Early mobile needed integration specialists for things that became SDKs
FDEs proliferate when technology is powerful but not yet interoperable. That's exactly where AI is today.
As the stack standardizes, the human translation layer compresses. Not gone, just normalized. The best companies in AI are already thinking about how to productize deployment workflows.
English
Fluidstack's high desirability is linked to their greater share of Anthropic exposure. Meanwhile, Crusoe's greater Stargate / OpenAI exposure has pulled them down on secondaries demand list: x.com/HarryStebbings…
TechCrunch@TechCrunch
AI datacenter startup Fluidstack in talks for $1B round at $18B valuation months after hitting $7.5B, says report techcrunch.com/2026/04/14/ai-…
English
Congratulations to @AndrewDai and the @ElorianAI team!
Andrew M. Dai@AndrewDai
After almost 12 years in Brain/DeepMind, I’ve finally decided to take the leap. My cofounders: @yinfeiy, Seth and I have kicked-off @ElorianAI. The first multimodal reasoning lab founded and led by former LLM pretraining, data and multimodal leads. youtu.be/YlvfNpOMeOY?si… (1/n)
English
It’s been a year. The last major model launch from $META was Llama 4 on April 5 2025. One of the biggest differences for this new Muse Spark model series: it’s a proprietary model that you have to pay META for, compared to the open-source Llama series.
Alexandr Wang@alexandr_wang
1/ today we're releasing muse spark, the first model from MSL. nine months ago we rebuilt our ai stack from scratch. new infrastructure, new architecture, new data pipelines. muse spark is the result of that work, and now it powers meta ai. 🧵
English
Despite the loss of a $200M DoW contract in March 2026, @AnthropicAI has now surpassed $30B of run rate revenue. What is behind this massive topline acceleration? 🧐
Everyone knows about their coding traction. But, Anthropic has also seen consumers flock to the Claude product after the company's DoW showdown. Whether this was intended or not, the company ended up trading $200M from the DoW for a few $ billion in run rate and consumer good will - not bad ROI...💸
Q1 2026 is likely the point where Anthropic's acceleration puts them at a larger reported run rate revenue scale than @OpenAI. But run rate revenue scale is just a point in time metric. The real deal to pay attention to is the slope of their growth 🚀 (Note: not a perfect apples-to-apples comparison due to different revenue share accounting)
Anthropic@AnthropicAI
Our run-rate revenue has surpassed $30 billion, up from $9 billion at the end of 2025, as demand for Claude continues to accelerate. This partnership gives us the compute to keep pace. Read more: anthropic.com/news/google-br…
English
Selina Wang retweetledi
I'm excited to announce we have achieved our Tier 1 mission success criteria and have begun gathering a tremendous amount of data on how this brand new spacecraft performs.
On March 30th, 13:17:08Z, the Gravitas spacecraft separated from the SpaceX Transporter-16 stack to begin its mission as one of the highest power free-flying satellites ever launched.
Immediately after separation, the spacecraft autonomously:
- Executed detumbling maneuvers
- Established two-way communications with the ground (on our very first ground station pass)
- Deployed its 20kW solar arrays
- Slewed to a safe and stable attitude to await further ground commands
These actions alone are a testament to the incredible work of our in-house engineering, software, and GNC teams to build a robust spacecraft.
Since then, our operations team completed all initial system activations and checkouts, confirming the vehicle is in a power positive and thermally stable state with no major anomalies observed at this time.
We completed this phase of the mission ahead of schedule. Next up we will be powering up and downlinking data for all payloads aboard the Gravitas spacecraft in support of our customers and partners while continuing to put the spacecraft through its paces.
As we noted ahead of launch: The goal of this mission is to experiment and push our systems to the limit to inform future missions. I look forward to sharing more on our successes and challenges as the mission proceeds.
Video of our satellite below; link to full T-16 webcast: x.com/SpaceX/status/…
English
A fantastic example of how companies with distribution reach can layer on AI products.
Ramp Labs@RampLabs
Today, we're launching Ramp Agent Cards. There's been no safe way for agents to spend money, until now. Ramp Agent Cards give agents the ability to spend, governed with real spend limits, merchant controls, and full visibility into every transaction.
English
Selina Wang retweetledi
I'm hiring some of the world's top AI researchers to join a new team I'm creating at Mercor, focused on building frontier benchmarks.
If you're an exceptional fit (or know someone else who is), dm me!
English
🚨People are underestimating the energy bottleneck for AI.
ERCOT alone is tracking ~205 GW of large load interconnection requests by 2030. ~70% are for data centers.
For context, ERCOT's entire current peak capacity is ~85 GW. The queue is exploding but the grid isn't.
🤯This is one state in one country.
English
Spend data from @tryramp shows declining labor marketplace spend and exponential AI model co spend.
Ara Kharazian@arakharazian
We have a definitive answer now: AI will affect the labor market. Starting with freelancers. In the first paper to use business-level data to track AI vs. labor. New paper from @tryramp finds: - Businesses are shifting spend from freelancers to AI. - More than half of the businesses using freelancers in 2022 have stopped entirely. - The companies that used to spend the most on freelancers shifted to AI the fastest. - 97% savings for the businesses that spent the most on freelance
English
🚨Massive funding round announced by Anthropic today. The company shared they've reached $14B+ of total RRR and Claude Code is at $2.5B+ of RRR.
1) How much of this growth is coming from Claude Code? By my outside in estimates, ~25% of the incremental RRR growth over the past 2 months is from Claude Code directly.
2) What is the remaining $4.5B of incremental Other RRR? I would speculate the vast majority of it is driven by other forms of coding usage on Opus 4.5 and 4.6
TLDR, coding remains the biggest driver of growth for Anthropic🤖
Anthropic@AnthropicAI
We’ve raised $30B in funding at a $380B post-money valuation. This investment will help us deepen our research, continue to innovate in products, and ensure we have the resources to power our infrastructure expansion as we make Claude available everywhere our customers are.
English
Awesome fireside with Unconventional AI hosted by @MollySOShea and @Playground_VC tonight. A few takeaways:
1. The biggest bottleneck to scaling AI is power ⚡️
2. They are making the bet that true intelligence will depend on an analogue chip architecture that mimics the human brain 🧠 and uses orders of magnitude less power
3. Their North Star KPI is joules per token ✨
4. CEO @NaveenGRao is jacked 💪🏼
English
I’ve gotten many questions about what happens to software developers as Claude models continue to improve. I think intelligence follows Jevons’ paradox, but some roles will simply disappear. Who’s starting the General Assembly for skilled trades?👀
Great data points from @benedictevans
English
Congratulations to @maltekosub and the @parloa team!
Parloa@parloa_ai
BREAKING: Just seven months since our Series C, we’re excited to announce our next big step: our $350 million Series D at a $3 billion valuation. 🚀 With this raise, led by @generalcatalyst, we’re accelerating our global growth, advancing our AI agent management platform (AMP), and launching the Parloa Promise – a new initiative aimed to redefine what’s meant by responsible AI. There is a clear signal. Customer patience for bad service has run out, and they’re leaving brands that don’t deliver on support promises. That’s why our mission is to eliminate these detached experiences and empower global enterprises to build meaningful relationships with their customers, turning every conversation into lasting loyalty and business value. 🔗Learn more here: parloa.com/parloa-series-…
English