Silicon Valley Daily

Silicon Valley Daily

Tech news du jour. Not accepting new members right now.

Dan Romero pfp

Ben Evans annual presentation is always good food for thought. https://www.ben-evans.com/presentations

4 replies

4 recasts

27 reactions

Dan Romero pfp

24h * 30 days = 720 a month = $27 an hour And that's not counting if you can have serial agents for the $20K a month plan.

2 replies

1 recast

14 reactions

Dan Romero pfp

Price discovery seems like a smart move (and traditionally most software has opted for lower price, greater volume). Also good narrative for fundraising. Question is how long does the premium last for with Grok, DeepSeek, Meta, etc.

2 replies

0 recast

13 reactions

Dan Romero pfp

“out of GPUs” 👀

4 replies

0 recast

11 reactions

Dan Romero pfp

Claude 3.7, Grok 3, Deep Research and Gemini updates all in a week Incredible amount of consumer surplus. Unclear how much will translate into value capture.

11 replies

6 recasts

66 reactions

Dan Romero pfp

No tweets from The Browser Company of New York for a month. Wonder what that means.

16 replies

1 recast

37 reactions

Dan Romero pfp

Good post on SV from Alex Danco https://open.substack.com/pub/danco/p/innovation-takes-magic-and-that-magic?r=2dg5&utm_medium=ios

3 replies

0 recast

10 reactions

Dan Romero pfp

Wild to think the US inherited so much of what makes it great from the UK—and how far the UK has fallen from its apex.

9 replies

13 recasts

64 reactions

Dan Romero pfp

Coinbase wins against the SEC but stock trades down. 1. This got priced in in early November. 2. Company is firing on all cylinders product-wise (from my outside perspective), but the new pillars like Base are still early from a revenue story (for a company that makes 6B+ a year) and market doesn't really understand the blockspace business 3. Coinbase is a index on overall crypto utility. The current price is an index on trading — both CEX and DEX — and a future weighting of increased competition (which seems more likely now). 4. The major drivers for the company will be: a. New breakout consumer use case(s) with crypto emerge b. Bitcoin moons Both will likely happen, but option A is the one they have more control over, esp. with Base.

6 replies

3 recasts

66 reactions

Dan Romero pfp

Reddit at $30B was not on my bingo card. 1. While traditionally their ad product was bad, they have been smart in leaning into the "[query] reddit" organic behavior. Other than Google and Amazon, one of the few at-scale websites that have consumer intent. Very good for ads. 2. LLMs are massively transformative for their business: a) data licensing to all the major players b) automatic translation for UGC advice / recommendations (see "[query] reddit"). 3. No meaningful competition from feed-based social networks. Missed opportunity for Quora.

16 replies

7 recasts

65 reactions

Zinger ↑ pfp

One of the most insane things about the Humane news is that all of the pins will basically become completely useless in 10 days since they were so stubborn about relying on their own services What a waste of time, talent, and money, reminds me of Magic Leap

9 replies

4 recasts

40 reactions

ted (not lasso) pfp

ted (not lasso)

wow, Humane getting acquired by HP for $116M. founders will focus on integrating into HP devices and systems. overpromised, undelivered. bad outcome imo.

10 replies

1 recast

37 reactions

Dan Romero pfp

everyone stopped using the vision pro after a week. what can we do to get people excited again? let's add the AI thing even less helpful than siri problem solved

17 replies

12 recasts

154 reactions

Nicholas Charriere pfp

Nicholas Charriere

This, then the response from Sam that Musk makes him sad and is insecure, is getting a little out of hand? Just billionaires fighting like kids in recess. Very kendrick and drake vibes

4 replies

1 recast

7 reactions

woj pfp

talked to an ai chad yesterday, asked him about deepseek his response: - it was bound to happen - he expects more hfts to publish their own LLMs and outperform openai etc by a mile research labs created the game but hft teams will master it due to: - better financial incentives - talent advantage great for competitiveness ans for the end users, most likely bearish openai/nvidia (at these prices)

9 replies

1 recast

35 reactions