Then why don’t you tell @hanera to stop wasting his time with Austin.
Save them, you or your kids can auction them later ! I actually wanted to have one (I did not think it was precious at that time !!)
Not at all, as I know the neighborhood very well. I am justifying or providing proof to you.
This is like hanera buying apple at 1998. When many here (in redfin forum) suggested sell, I firmly stood strong as I believed in that location. Even during 2008-2011, this location did not fall significantly. I turned this into rental in 2011 (the worst period of bay area real estate) confidently.
I was confident on falling TSLA (bought 2728 shares at $145-153 sold at $258) and Falling SHOP, took big stake challenging my skills.
When I am clear and confident, I am not disturbed by any statements/discussions/reports (CITRON kind of).
It is hanera’s forte, he did his research and making investments based on his study. In fact, I feel He is sharing his knowledge with us.
It does not mean Austin is great, neither SV is great. It is individual’s idea. We get benefit out of his knowledge sharing (IMO, That is the key for this forum).
I think most areas in SV will top out at the 2-3M price range. They may still go higher but slowly. Neighboring towns will see price squeezed higher as buyers flee the more expensive towns. We see this dynamics play out in Sunnyvale and now Santa Clara.
I don’t see Cupertino prices going up 7% YoY from the current price point. That’s saying average Cupertino houses will be worth 5-6M in 2027.
Ok well East Bay is my forte and I am sharing my knowledge with you guys. Anyone who dares to question that location will get his ass kicked
No doubt, no expectation to see 7% YOY for CU,past results does not mean future returns assured, but as you said you can expect low price area can pickup at this rate. It is very natural.
My main content is not about Cupertino (even though I showed CU posting), but to say Bay Area will have bright future for multi decades even after programming is done by automation. I see at least 50 more years growth.
Yes, many real estate people are area specialized. After your postings, I reviewed many Hayward locations next to Fremont which may pick up better appreciation.
In fact, after your postings and elt1 posting, I started looking opportunities in East Palo Alto, but homes are skyrocketing there too.
For investment purpose, Low cost and cash flow are the key.
When manch was thinking about 529 plan and how to fund, I was keeping Oakland and Hayward in mind to get a low cost rental to generate $1500 free cash flow and invest returns it in 529 plan.
Actually, even the model creation is being automated. Google has shown that automated techniques can create neural network architectures which exceed the best architectures created by a person (see link below). This is often called “AutoML”.
I’m excited because the demand for ML compute (both training and inference) is insatiable. If the world had 1000x more ML compute, it would still not be enough. This is going to be another renaissance for chip companies, as they shift towards ML specific processors.
Few realize many keypunch cards are made right here on Bascom in Campbell, CA
In July 1960, punched card manufacturing was moved from San Jose to a new facility in Campbell, Calif. They need a skilled technocrat to close the doors. Same will happen to many companies SWE or not.
15 mil thickness. Great to use it as a book marker and a shim to validate spark plug gap.
Which chip companies are you thinking of?
Are NVDA, AMBA, CRUS and SYNA any good?
Too lazy to reply to… basically, 1 article says no more programmers… another who own RE in East Bay agreed… hence the logical deduction is since East Bay even though might not have that many programmers, but many businesses there serve SF & SV, so it would be affected too. So that some1 who own RE in East Bay and agree that there would be no more programmers should sell RE in East Bay now even though it is not directly affected as in SF & SV.
Having said above, I use SWEs and not programmers, is meant to be cheeky, both are not synonymous. I think the demand of SWEs would continue to be strong for a few decades… their jobs are not programming only… moving up the higher value aspects of software engineering.
All chip makers will have to adapt. NVDA got there first, not sure if they can stay on top. The startups in this space are looking a bit bubbly. Cerebras was valued at 1B with no product.
Probably miss the run already… have to wait for some consolidations to see which are the possible winners… usually Mr Market would indiscriminately sell when he gets bearish about the industry which give us the opportunity to grab the winners.
Is AI chip primarily for data center? It seems a low volume business. Google can develop its own AI chip to get an advantage over other competitors.
I’m skeptical of AI chip startups. Does Nvidia get its major revenue from mon-AI chips? It’s GPU has a larger market outside of AI
Wait, even though I agreed that programmers are endangered, doesn’t mean I believe Bay Area real estate value will drop with that.
Good talk by Peter Norvig on how programming with ML is different from traditional programming. At the end he went over the difficulties of ML.
I am starting to think more about AI as I realized, belatedly, it’s a massive paradigm shift. If software programs are more human like, you don’t “program” them per se. Just like you don’t send your kids to school to be programmed. Data orientation is orthogonal to control orientation.
Right! We are in the middle of a hardware renaissance. The most exciting tech company in the last 2 years is Nvidia. This week Intel stock broke into $40, the highest since the dot-com bust.
At a very high level ML is throwing lots and lots of hardware at lots and lots of data and answers magically show up. Hardware does the work, software taking a back seat. 10 years ago it was “software eating the world”. Now it’s “hardware+data eating the world”. 10 years ago no VC will touch chip startups. Remember Transmeta? Now it’s one of the hottest areas.
Actually, is not too late to invest in chip companies, NVDA is winning today but the field is very wide so could have many winners or at least possible buyouts of smaller companies by big guys like NVDA, INTEL or even AAPL and GOOG are possible i.e. many opportunities to make money still.
Given that you’re a chip guy, can you share with us your recommendations of potentials… I won’t hold you to it if they don’t work out. So far, I’m betting on NVDA (general purpose AI), AMBA (Computer Vision chip), CRUS (AI version of audio chip) and SYNA (AI version of far-field voice chip). I’m not a chip guy, selection is based on reading articles on the web Investment is only modest because I just don’t know much, need some1 who know to provide more depth.
Btw, Santa Clara city is on my RE buy list before I decide on Austin. The renaissance of semi would last for awhile but I heard that China is pretty fast on this trend. So I hesitated as not sure whether semi companies in SV can compete with China. In fact, China has jumped into AI way faster than most nations e.g. Singapore, Singapore is now desperate, too slow to adapt to the new AI trend.
AI is so important that UAE appointed a Minister of AI.
The only AI related investment I have made so far is Micron. Regardless of whether it’s Nvidia, AMD, Intel or some new startups they all need RAM. Supply is controlled by 3 companies only.
I am still not sure about Nvidia. On the cloud side google already has its own chip. Facebook is working with intel. Amazon for sure is cooking something. Intel has been on buying spree acquiring AI startups and has now 5 or 6 different offerings. Lots of action too in self driving cars from many players including startups. Nvidia’s super high PE gives me pause. I am not sure it has the moats to justify it.
A very good write-up of AI’s implications on software, written by Tesla’s AI director.
Read the whole thing. Especially good are the benefits section. Here’s the part that drives the hardware renaissance:
The benefits of Software 2.0
Why should we prefer to port complex programs into Software 2.0? Clearly, one easy answer is that they work better in practice. However, there are a lot of other convenient reasons to prefer this stack. Let’s take a look at some of the benefits of Software 2.0 (think: a ConvNet) compared to Software 1.0 (think: a production-level C++ code base). Software 2.0 is:
Computationally homogeneous. A typical neural network is, to the first order, made up of a sandwich of only two operations: matrix multiplication and thresholding at zero (ReLU). Compare that with the instruction set of classical software, which is significantly more heterogenous and complex. Because you only have to provide Software 1.0 implementation for a small number of the core computational primitives (e.g. matrix multiply), it is much easier to make various correctness/performance guarantees.
Simple to bake into silicon. As a corollary, since the instruction set of a neural network is relatively small, it is significantly easier to implement these networks much closer to silicon, e.g. with custom ASICs, neuromorphic chips, and so on. The world will change when low-powered intelligence becomes pervasive around us. E.g., small, inexpensive chips could come with a pretrained ConvNet, a speech recognizer, and a WaveNet speech synthesis network all integrated in a small protobrain that you can attach to anything.
The software layer is much simplified, which then lends itself to hardware acceleration. The companies that makes the following happen will have 1T market caps: small, inexpensive chips could come with a pretrained ConvNet, a speech recognizer, and a WaveNet speech synthesis network all integrated in a small protobrain that you can attach to anything.