That only summarizes the first two paragraphs. The article goes on about VC funding and office expansions among other things.
The flow of venture capital dollars into AI and machine learning companies in San Francisco hit new highs this year, with start-ups raising $18.5 billion in the first quarter — about 82 percent of U.S. investments in the segment
There's no denying the fact that SF is the place for AI startups.
One difference I would guess is scale. In SF there are AI meetups every day, and the big ones on weekends have literally hundreds of developers actively writing code all day. There are typically at least 2 events with >100 developers every Saturday and probably 20-30 of various smaller sizes.
Cool, so there are AI meetups, which are finite sessions inside random cafes and businesses. But what about literally everything outside those meetups, like the horrifying human experience involved in the commute/walk/bike to and from said meetups?
You act like the meetups are happening in the Tenderloin.
I don't care to attend an AI meetup, but I just googled "sf ai meetup" as a thought experiment. First result is 3rd and mission. A very tame block. I could get on the N Judah and wind up near there and not face any of this "horrifying human experience". I do that same trip with my children since they were babies. It's no big deal. I would question the sanity of anyone who is concerned about that trip.
> horrifying human experience involved in the commute/walk/bike
Have you been to SF? 95% of my trips around SF are far better than the average soul sucking drive around the rest of the country. Most of it has lovely architecture, beautiful parks, and great spots to stop into. For sure, there can be more extremes in the other direction, but imo they aren't that common and don't much bring down how much better the average is. Plus they're concentrated in tenderloin / civic center and most people will avoid that area to reduce ever encountering crazy stuff.
Unfortunately this boom is way different than the previously innovation driven tech booms. There’s no competitive advantage in writing openai wrappers. Any wealthy city can be a hub for it.
What kind of innovation had competitive advantages in the previous booms? Did building Facebook or Twitter have any inherent technical or competitive moats, or could anyone else also build those at the time?
That inspired me to copy your comment as a prompt in ChatGPT:
Artificial intelligence (AI) has the potential to impact various aspects of the commercial real estate industry, but it may not have the ability to save or prevent a commercial real estate bubble on its own. The state of the commercial real estate market is influenced by a complex interplay of economic, financial, and market forces, and the rise or fall of a bubble is typically tied to broader economic conditions.
Here are some ways in which AI can be relevant to the commercial real estate industry:
1. Data Analysis: AI can help analyze large volumes of real estate data, such as property values, rental rates, and market trends, to provide insights to investors, developers, and property managers. This data analysis can inform decision-making and help identify investment opportunities.
2. Predictive Analytics: AI algorithms can be used to make predictions about future trends in the real estate market, which can aid in making informed investment decisions and risk assessments.
3. Automation: AI can automate various aspects of property management and maintenance, potentially reducing operational costs and increasing efficiency for property owners.
4. Risk Assessment: AI can assist in assessing the creditworthiness of tenants and help in the underwriting process for commercial real estate loans.
5. Virtual Tours and Visualization: AI-driven technologies can enhance property marketing by creating virtual property tours and realistic visualizations of properties.
While AI can provide valuable tools and insights, it cannot single-handedly save a commercial real estate bubble. The rise or fall of a real estate bubble typically involves a combination of factors such as interest rates, economic conditions, supply and demand dynamics, and investor sentiment.
To address issues related to a potential real estate bubble, it often requires a combination of government policy, prudent lending practices, and regulatory oversight. AI can be a part of the solution, but it is just one component of a complex ecosystem in the real estate industry.
Is there value in moving there for an individual who’s an employed swe, but trying to break into AI? I’d love to work for some of these startups but it seems difficult to get past the resume screen these days, and I don’t think I have that weak of a resume. If moving there would make a big impact I’d consider it, but I’m not sure where exactly in SF, and how to connect to the community while maintaining my current job remotely
> Is there value in moving there for an individual who’s an employed swe, but trying to break into AI?
Yes. If you’ve never experienced the value of an industrial cluster, it’s exhilarating. You still have to hustle. But the payoff on a single day well executed is exponential; meetings beget introductions and actions are decided on handshakes.
So if I move there I should maximize meetups? How does one filter meetups? I only ask because I do see a large quantity of meetups, but having never been there before I’m not sure what to expect, and I’m considering a short term (30-60 days) trip to determine if moving there is a good idea for my career or if I should just stick to remote work in the midwest
I think the one thing we can all agree on is that we're actually excited about something in tech again. Like for real, not in the "oh thats interesting now how can we make money" kind of way. It's been a long time, well over a decade since that's happened. The energy of SF in the late 2000s with the mobile boom was electric. Would be wonderful to experience that again.
> You didn't feel it two years ago when crypto and Web3 were going to fundamentally change everything?
That was the charlatan crap that took over in the last 10 years. Once upon a time, people actually cared about tech for it's own sake, and spent time geeking out and having meetups for the fun of it. Not everything was a "networking opportunity" to grift into a 6 figure PM role. I'd say 2014ish was about the time those things started becoming >50% nontechnicals and recruiters, which was about the same time that every last drop of real culture in that city made an exodus to the east bay.
I think 99% of people who were into that just wanted to get rich quick when the crypto prices went apeshit. Recent AI feels like crazy sci-fi stuff on its own merit, whether it makes money or not.
>Yeah but this time the tech is based on stealing people's intellectual property. Hopefully this theft ends soon.
It isn't necessarily or fundamentally, though. Public data was just the most accessible training corpus for the POC phase of adoption. Getty just released their first licensed model, and I suspect things like that will pick up steam as the tech has rapidly commoditized.
That being said we will probably end up with a supreme court ruling on this in the next five years, and I strongly believe it will come down on the side of fair use.
Technically it is only taking gradients from the training text, not the text itself. And those gradients are all added up so the contribution of any one of them is diluted to almost nothing.
You cannot expect to be paid the same amount of money by the same size audience when the supply increases dramatically. Recording industry got this lesson 20 years back.
What if what you can build based on humanity's collective "intellectual property" is worth the discomfort of not "owning" the work you release into the world?
Having energy and money go to something that is not crypto is so refreshing. I’m having conversations about how to use technology to make things cheaper and easier again!
Happy to see SF revitalized by the work on these cool new challenges. Solving the hard problem of replacing all office workers with AI really requires in-person collaboration.
Readit News
Summary: Meetups, especially AI meetups, are back in SF.
Love to see it, but we're also seeing the same thing happen here in DC with the meetups I run, so it's not especially newsworthy?
I don't care to attend an AI meetup, but I just googled "sf ai meetup" as a thought experiment. First result is 3rd and mission. A very tame block. I could get on the N Judah and wind up near there and not face any of this "horrifying human experience". I do that same trip with my children since they were babies. It's no big deal. I would question the sanity of anyone who is concerned about that trip.
Have you been to SF? 95% of my trips around SF are far better than the average soul sucking drive around the rest of the country. Most of it has lovely architecture, beautiful parks, and great spots to stop into. For sure, there can be more extremes in the other direction, but imo they aren't that common and don't much bring down how much better the average is. Plus they're concentrated in tenderloin / civic center and most people will avoid that area to reduce ever encountering crazy stuff.
So yeah, not much.
Dead Comment
Artificial intelligence (AI) has the potential to impact various aspects of the commercial real estate industry, but it may not have the ability to save or prevent a commercial real estate bubble on its own. The state of the commercial real estate market is influenced by a complex interplay of economic, financial, and market forces, and the rise or fall of a bubble is typically tied to broader economic conditions.
Here are some ways in which AI can be relevant to the commercial real estate industry:
1. Data Analysis: AI can help analyze large volumes of real estate data, such as property values, rental rates, and market trends, to provide insights to investors, developers, and property managers. This data analysis can inform decision-making and help identify investment opportunities.
2. Predictive Analytics: AI algorithms can be used to make predictions about future trends in the real estate market, which can aid in making informed investment decisions and risk assessments.
3. Automation: AI can automate various aspects of property management and maintenance, potentially reducing operational costs and increasing efficiency for property owners.
4. Risk Assessment: AI can assist in assessing the creditworthiness of tenants and help in the underwriting process for commercial real estate loans.
5. Virtual Tours and Visualization: AI-driven technologies can enhance property marketing by creating virtual property tours and realistic visualizations of properties.
While AI can provide valuable tools and insights, it cannot single-handedly save a commercial real estate bubble. The rise or fall of a real estate bubble typically involves a combination of factors such as interest rates, economic conditions, supply and demand dynamics, and investor sentiment.
To address issues related to a potential real estate bubble, it often requires a combination of government policy, prudent lending practices, and regulatory oversight. AI can be a part of the solution, but it is just one component of a complex ecosystem in the real estate industry.
Yes. If you’ve never experienced the value of an industrial cluster, it’s exhilarating. You still have to hustle. But the payoff on a single day well executed is exponential; meetings beget introductions and actions are decided on handshakes.
/s (sorta..)
That was the charlatan crap that took over in the last 10 years. Once upon a time, people actually cared about tech for it's own sake, and spent time geeking out and having meetups for the fun of it. Not everything was a "networking opportunity" to grift into a 6 figure PM role. I'd say 2014ish was about the time those things started becoming >50% nontechnicals and recruiters, which was about the same time that every last drop of real culture in that city made an exodus to the east bay.
It isn't necessarily or fundamentally, though. Public data was just the most accessible training corpus for the POC phase of adoption. Getty just released their first licensed model, and I suspect things like that will pick up steam as the tech has rapidly commoditized.
That being said we will probably end up with a supreme court ruling on this in the next five years, and I strongly believe it will come down on the side of fair use.
Deleted Comment