3 ms·

Let's see what's left in Alphabet that isn't Google. There's Calico, CapitalG, Fiber, GV, Intrinsic, Isomorphic, Verily, Waymo, Wing, X. Nest and DeepMind were

by dekhn 2y ago

Let's see what's left in Alphabet that isn't Google. There's Calico, CapitalG, Fiber, GV, Intrinsic, Isomorphic, Verily, Waymo, Wing, X. Nest and DeepMind were absorbed into Google, and many of the previous Alphabet mini-companies are already gone (Makani, etc). My personal expectation is that Calico will eventually wind down, as their research has been pretty unexceptional and I don't think they've managed to find any real revenue sources. Intrinsic is probably on borrowed time, looking for revenue, and if they don't get significant amounts in the next two years, will be shut down. Waymo isn't profitable yet but it looks like the execs have decided to give it a very, very long runway, and they are the technical leader in no-driver cars. It's still at risk if they don't find a good revenue model. Wing has no real revenue to support their costs, and drone deliveries just aren't going to be a great business, so I expect it to shut down in the next two years. Fiber- it's hard to tell why Google has allowed this to continue for so long. The only thing I can imagine is that they are somehow using the usage data to generate revenue or ML training data. Intrinsic is really new so they probably have a couple years runway, but pharma sales aren't a great business to be in if your only product is a protein structure predictor and binding predictor. I work in this field and I still see this as an unrealistic bet on Alphabet's part. That leaves Verily. It's clear Alphabet has downsized it significantly and scoped the projects down significantly. They have turned away from every single early idea and pivoted to a few remaining areas where they have some strength. I work in this field and it's amazing to me they keep it around- so I expect that unless they sell it in the next year or two, it will be shut down. CapitalG, GV, and Google's real estate ventures will probably all continue because they have generated a fair amount of revenue. Google Cloud is part of Google, not Alphabet, which IMHO is an insult to the people who run it. It could be spun off as an all-in Alphabet company but it's so closely tied to the rest of the Google infrastructure it will never live on its own. Truly impressive how little Alphabet has achieved.

zamadatix 2y ago
Google Fiber seems like a reasonable business, I'm just more surprised they haven't went and sold it off already.
toast0 2y ago
> Fiber- it's hard to tell why Google has allowed this to continue for so long. The only thing I can imagine is that they are somehow using the usage data to generate revenue or ML training data. I'd guess they signed long term agreements with the municipalities where they offer service and they can't find a good way out. That or they have acccess to some resource related to being a residential ISP that they want to use for other things (cable TV feeds?). Fiber sort of fulfilled its mission of getting more people connected at high speeds; the problem is the best way to do that is to annouce a build out in 20 cities, and look busy for 18 months while the telco and cableco upgrade their existing customers to something similar. I don't think you can do that more than once though. Actually building a network isn't fiscally responsible, since they won't see many takers when the incumbents can upgrade faster than Google can overbuild and most people are fine with ~50/5 service anyway.
- whaleofatw2022 2y ago
  > That or they have acccess to some resource related to being a residential ISP that they want to use for other things (cable TV feeds?). I wonder if it is also part of some sort of peering or 'insurance policy' against large isps/etc (i.e. net neutrality)
l33tman 2y ago
As someone who previously did research in MD and simulating protein kinematics and binding, may I ask what you who are in the field today think about the complement so to speak - doing synthetic structures from 3D templates? Like David Baker's lab (I haven't been keeping track, to me it seemed like he was the only one in the world going all-in on this).
- dekhn 2y ago
  I think MD is a big waste of time and resources for nearly all problems, even free energy prediction. Maybe useful for studying kinetics of sub forms, ala Pande. Baker and deep mind lead the field with many ankle biters nipping at their heels. It's al proceeding roughly as I expected when I did casp in 03-04
xnx 2y ago
Waymo has a huge potential totally addressable market. The possibility to earn money on every mile driven. Hardware costs are high now, but are on a very favorable curve for cost and capability. Potential for Wing also seems good. Any delivery under 5 pounds an under 10 miles is going to be much faster/cheaper by air.

by dekhn 2y ago

zamadatix 2y ago

Google Fiber seems like a reasonable business, I'm just more surprised they haven't went and sold it off already.

toast0 2y ago

> Fiber- it's hard to tell why Google has allowed this to continue for so long. The only thing I can imagine is that they are somehow using the usage data to generate revenue or ML training data. I'd guess they signed long term agreements with the municipalities where they offer service and they can't find a good way out. That or they have acccess to some resource related to being a residential ISP that they want to use for other things (cable TV feeds?). Fiber sort of fulfilled its mission of getting more people connected at high speeds; the problem is the best way to do that is to annouce a build out in 20 cities, and look busy for 18 months while the telco and cableco upgrade their existing customers to something similar. I don't think you can do that more than once though. Actually building a network isn't fiscally responsible, since they won't see many takers when the incumbents can upgrade faster than Google can overbuild and most people are fine with ~50/5 service anyway.

whaleofatw2022 2y ago
> That or they have acccess to some resource related to being a residential ISP that they want to use for other things (cable TV feeds?). I wonder if it is also part of some sort of peering or 'insurance policy' against large isps/etc (i.e. net neutrality)

l33tman 2y ago

As someone who previously did research in MD and simulating protein kinematics and binding, may I ask what you who are in the field today think about the complement so to speak - doing synthetic structures from 3D templates? Like David Baker's lab (I haven't been keeping track, to me it seemed like he was the only one in the world going all-in on this).

dekhn 2y ago
I think MD is a big waste of time and resources for nearly all problems, even free energy prediction. Maybe useful for studying kinetics of sub forms, ala Pande. Baker and deep mind lead the field with many ankle biters nipping at their heels. It's al proceeding roughly as I expected when I did casp in 03-04

xnx 2y ago

Waymo has a huge potential totally addressable market. The possibility to earn money on every mile driven. Hardware costs are high now, but are on a very favorable curve for cost and capability. Potential for Wing also seems good. Any delivery under 5 pounds an under 10 miles is going to be much faster/cheaper by air.