A
@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: https://t.co/XgSjL0And7
5.3K
185
2.9M
@AnthropicAIposted on X
We've signed an agreement with Google and Broadcom for multiple gigawatts of next-generation TPU capacity, coming online starting in 2027, to train and serve frontier Claude models.
View original tweet on X →
Community Sentiment Analysis
Real-time analysis of public opinion and engagement
Sentiment Distribution
74% Engaged
51% Positive
23% Negative
Positive
51%
Negative
23%
Neutral
26%
Key Takeaways
What the community is saying — both sides
Supporting
1
Compute is the moat
Many replies treat this as proof that the AI race has shifted from model design to raw infrastructure — "data centers are the new factories" and securing gigawatts locks a competitive lead.
2
Big, long-term hedge on scale
Committing multi-gigawatt TPU capacity starting 2027 is read as a deliberate bet that scaling laws will keep paying off — the jump to a $30B run-rate is cited as validation so far.
3
Silicon diversification = risk management
Commenters note Anthropic is running a portfolio approach (TPU + Trainium + NVIDIA) to avoid single-vendor choke points and keep unit economics sane.
4
Market and vendor implications
The deal is seen as bullish for Google and Broadcom ($GOOGL, $AVGO), potentially undercutting Nvidia’s leverage ("the Nvidia tax") and reshuffling winners in the supply chain.
5
Power & physical constraints matter
Multiple replies emphasize that the true bottleneck may be the grid, substations and power procurement — this is "utility-scale compute" that requires real-world infrastructure buildout.
6
Barriers to entry will harden
Locking multi-year compute reservations is viewed as industry crystallization: access to silicon becomes an admission price, concentrating advantage among a few chip designers and cloud providers.
7
Immediate product and user upside
Users expect higher throughput, fewer rate limits and faster model release cadence — many replies hope the capacity will translate into more powerful, less throttled Claude features.
Opposing
1
Multi‑gigawatt power demand is alarming:
Many replies treat “gigawatts” as a practical red flag — concerns about huge power bills, where the electricity will come from, and the idea that AI is becoming energy infrastructure, not just software.
2
Compute supply = strategic lock‑in:
People argue this is a power play, not a product race — firms are reserving future compute capacity, creating barriers that favor a few deep‑pocketed players.
3
Cloud partner conflicts of interest:
Several callers point to the oddity of using an investor/cloud provider/competitor (Google) for massive capacity — framed as a potential conflict of interest and risky dependence.
4
Token drains and rate limits are driving users away:
Repeated complaints about drastic token consumption, 429 errors, and tight quotas — many say usability is crippled and users are cancelling or threatening to leave.
5
Product quality and feature breakdowns matter more than scale:
Critics note Claude’s recent flakiness, missing or crippled features (OpenClaw), and canned responses — scaling compute won’t fix a product that users find unreliable or dumb.
6
Pricing and business model skepticism:
Comments range from “overcharging” to predictions of bankruptcy; some highlight that the advertised scale may just mean a bigger bill with diminishing consumer value.
7
Calls for different hardware and competition:
Users suggest alternatives (AMD, Cerebras), point to rivals running well on midrange phones/GPU, and urge focus on efficient architectures instead of brute‑force TPUs.
8
Hostile and political backlash:
Threads contain angry, conspiratorial, and abusive replies — accusations of ties to militarized tech, applause for competitors, and vindictive calls from banned or disgruntled users.
9
Sarcastic dismissal and existential fear:
A mix of jokey mockery (“buy candles,” “toaster’s got more personality”) and genuine AGI/Skynet worries — some see the scale announcement as either laughable or frightening.
Top Reactions
Most popular replies, ranked by engagement
O
@OfficialLoganK
TPU’s are truly incredible
518
12
14.7K
B
@brad3141592653
my GOD we are about to see another post like this again
146
0
5.0K
B
@BitcoinAIGuy
where u gonna get the power?
76
10
5.6K
C
@CausalEngineer
Google is backing open-source Gemma while also expanding Anthropic’s access to TPUs and Google Cloud. That level of platform leverage is remarkable.
64
0
6.8K
O
@OSherikar
Please work on the drastic drain of tokens, its very important we are just wasting the tokens coz of your bs
29
1
3.5K
Report an Issue
Found something wrong with this article? Let us know and we'll look into it.