ravst3s
Nat gas is preferred for AI DCs because it has faster time-to-market, doesn't have the intermittency issues. Training on solar + storage is an issue because of network synchronization.
Every GW of Blackwell generates more revenue than the entire gaming business does in 1 year.
These two products dominate every non-technical industry. I think MSFT makes mediocre software. But as long as those two gravy trains exist, they'll never lose.
New Intel financials: https://www.intc.com/filings-reports/all-sec-filings/content...
Old Intel financials: https://d1io3yog0oux5.cloudfront.net/_9bece0e2e2052ad835960e...
"Revenue was $18.9 billion down $8.6 billion from 2022. Internal revenue was $18.0 billion, down $9.1 billion driven by lower intersegment volume. External revenue was $953 million, up $479 million from 2022, driven by higher packaging revenue"
https://www.intc.com/filings-reports/all-sec-filings/content...
Twitter has 360m MAUs, 225m DAUs globally. That's less than what Facebook does in Europe.
"By the numbers: Engagement metrics are down across the board.
App downloads fell roughly 38% globally between October 2022 and September 2023, according to Sensor Tower estimates. In the U.S., mobile app downloads fell 57% in the same time period. Data from Data.AI, another app tracking firm, shows similar trends.
Usage has also decreased, with monthly active Android users falling 14.8% globally and 17.8% with mobile users in the U.S. year-over-year for the month of September, per SimilarWeb.
Average time spent, daily per user, fell 2% year-over-year globally in the third quarter of 2023. Sessions dipped 4% in that same time period, per Sensor Tower. User churn, or users who stop using the app, increased more than 30% year-over-year as of September 2023, per Sensor Tower.
Web traffic was down 7% globally and 11.6% in the U.S. for the first nine months of 2023 compared to the same period in 2022, per SimilarWeb."
Twitter has always been a power-user platform, but lacks mass appeal.
MAUs are down 15-20% YoY, gross churn is up 30%, but time spent is only down 2%.
Nothing besides $ prevented their competitors from developing their own GPUs and software stack comparable to CUDA.
Investing in China was already a thorny issue, you cannot trust the government at all as a foreign investor and there's 0 transparency. Now, there's 0 growth as well.
The last part is the most important as it will cap any thought of entering China again. If the government stimulates the economy massively, investors will return for a short period.
Thrive was about to buy employee shares at a $86 bn valuation. The Information said that those units had 12x since 2021.
https://www.theinformation.com/articles/thrive-capital-to-le...
Iirc, the NP structure was implemented to attract top AI talent from FAANG. Then they needed investors to fund the infrastructure and hence gave the employees shares or profit units (whatever the hell that is). The NP now shields MSFT from regulatory issues.
I do wonder how many of those employees would actually go to MSFT. It feels more like a gambit to get Altman back in since they were about to cash out with the tender offer.
I'm sure some of those employees were easily going to make $10m+ in the sale. That's a pretty great motivation tool.
Overall, I do agree with you. The board could not justify their capricious decision making and refused to elaborate. They should've brought him back on Sunday instead of mucking around. OpenAI existing is a good thing.
Bing chat sucks now. All it does is generate SERPs instead of encapsulating information.
Investing in RL is not a bad idea if it can drive significant top-line growth. However, unless RL turns into one of the best businesses on earth (like FoA), it will never delivered a positive ROI since its inception.
US debt is similar in that it's invested capital, but it does provide a positive contribution to GDP. If the US stops growing, our debt will become a massive issue.
Ironically, Google doesn't care about their shareholders and quarters like other tech companies. They don't provide a guidance, their earnings calls have little substance, and their disclosure is atrocious.
These companies are aiming to have installed bases in the hundreds of millions if not billions.
Active gaming will be a nice use case, but will not be the reason 40-50m people buy a headset every year.
Financially, RL is the worst business on earth. No other business has lost this much money in such a short period of time.
If you're gaming for 3-4 hours, watching a movie, or using the device for productivity you're going to be seated or standing still.
Also, Vision Pro can be a mobile device, albeit for <2 hrs. Imo, Gaming will not be the primary use case for these devices.
Although Meta will sell more units in the short-term due to Apple's pricing, I do think they will have a tough time catching Apple on the hardware. The eye-tracking, <12ms image processing and display, and the M2 are things that Meta is well behind on.
Zuck mentioned that activity and "doing things" are Meta advantages, but people don't want to exercise while gaming or consuming entertainment.