Does anybody else have the feeling "Microsoft is back"?
With GitHub and OpenAI the have put their finger on two pretty significant new projects.
What new things has Google come up with over the last 5 years?
With GitHub and OpenAI the have put their finger on two pretty significant new projects.
What new things has Google come up with over the last 5 years?
Google Brain, and by extension the Transformer model (~which is the T in GPT~ my bad, it's Generative PreTraining, not General Purpose Transformer), and TensorFlow.
Waymo, which is rolling out robotaxis in a few places.
Not sure which department or subsidiary is responsible for their TPUs, the Sycamore 53-qbit quantum processor, or LaMDA.
Waymo, ok, I don't live in that area. How can I try any of the other projects you listed?
Not sure if OpenAI and with the release of chatgpt has fully implemented their own call for responsible ai
pip install tensorflow
https://cloud.google.com/tpu/docs/tpus
Most of what I listed is groundbreaking research, not consumer stuff. For the consumer stuff, try e.g. searching images by description in the default Android app, or Google Translate, or the automated transcriptions on YouTube. Apple also does similar things, as stuff like that is what consumer AI looks like these days.
But I have an opinion on Google translate. From my experience, it is not clearly leading in comparison to other tools by smaller companies.
If Google is leading in AI, why don't they have any clearly leading consumer products in this area?
For machine translation, they used to be clearly leading, then deepl arrived and I think they are roughly on par (or maybe deepl is better, but also has fewer languages). Do you have other companies in mind?
As mentioned above, in research they have some great results, translating those to products is not always straightforward of course - or even possible.
It is true, however, that TensorFlow is now considered outdated, with pretty much all the community switching to PyTorch (originally developed by Facebook, now handed over to the Linux foundation). But indeed these are not really products (i.e. they don't bring in the money they cost, by far), more like open source projects.
Waymo so-far has been, until the last few months, unable to make regulators happy. Even then, the tech doesn't actually solve many of the original goals. Robotaxi's do not improve traffic, their solution is not environmentally friendly, and the payoff is still unclear, it is an uphill battle. The division loses incredible amounts of money, all while having trouble making good mindshare and when it does, for the wrong reasons.
Tensor is a great contribution...and don't get me wrong, I love OSS...but you can't just make an open framework and say 'Boom, product!' At most it makes for some new avenues for Web Services but not in a wide way that is highly profitable. In this regard, Microsoft's 1B investment in OpenAI and then just capitalizing on it is hilariously good value.
DeepMind is a news article system, sadly, despite incredible work by their team, Google has almost made it the butt of the joke: "We used Deepmind to personalize play store ads!". One day they might find a way to monetize it effectively, so far, it's kinda just like Google's IBM Watson problem: Market it to solve big data problems, actual impact ends up being unclear.
I personally don’t think there have been any amazing consumer tech breakthroughs by any big tech in the last 5 years.
https://scholar.google.com/scholar?q=%22language+model%22+&h...
Specifically by Google Brain, which is really quite different from the rest of the company. It's one of the top AI research labs in the entire world.
They're well into the "extend" stage for Linux. Trying to corner and ultimately monopolise the console gaming market.
Now they want to monopolise LLMs as much as possible.
Not sure this is a good thing
God help us when they get a near monopolistic position in video gaming and large-scale AI.
As the kids nowadays would joke, "Gates and Ballmer are living in their heads rent-free" except it's been for the past 25-30 years.
They have renewed their grip on several audiences.
Roblox by contrary runs on what amounts to a "financialize every individual interaction" model. That just becomes tiring for children after a while.
Kids are encouraged to spend as much money as possible on lots of bite-sized microgames, all existing on the "Roblox" platform, whilst being encouraged to use Roblox' own (proprietary black box, not reusable outside of Roblox btw) microgame creator to create more games, which they will alledgedly get paid for by the other kids buying those games and it can "be paid out in real money". That sounds got on paper, but in practice it's basically company scrip[0], since Roblox already takes a cut out of every internal currency purchase of a microgame, but once you try to convert it to real cash, you need to have an absurd amount of profit from your microgames first and they take *another cut* when you try to pay out. It's a model basically designed to keep every dollar converted to robux as robux.
At some point either mommy catches on that their kid is spending an awful lot of money on Robux/the kid gets frustrated that they're not getting their actual payouts because they want to say, buy a real world thing that you can't get with Robux and gets hit with the "earn 1k$ worth in Robux or we don't pay out" wall.
windows 11 seems like an intentional disaster where basif features are misaing because microsoft decided you didnt need to them. office seems to be getting worse and is now solely about thay subscription money.
and many of microsofts acquisitions just languish slowly - linkedin etc.
google can buy conpanies too. meh.
In fairness, this is true of most major companies in the past 20 years. Apple acquired PA Semi to build their A-Series chips, Google just straight up bought the Android OS. Facebook bought Oculus and renamed itself, Tesla's first 3 years of ASDS was just Mobileye's tech and Tesla overselling it's abilities.
Windows 11 being a 'disaster' is a vast overstatement of issues. Consumers widely seem to be fine with it by the numbers, and Microsoft found more ways to generate profits. Sadly they are a business. Adding ChatGPT to windows would wildly change the context of the product and would probably be a massive blow to Apple as ChatGPT is a real "I want that" feature if they market it to wide audiences, at the level of the first laser printer or such. A product where you don't have to know anything about how it works to know that the output is wanted.
Linkedin makes money, provides direct information of practices, and allows them to own a quiet corner of the majority of US user's life. Just owning and sitting on it is a great investment, even if you forget the whole LinkedIn learning/Certs thing they are doing to own the education corner.
B) Microsoft did not create GitHub or OpenAI.
C) Both Google and Microsoft have many significant research and development outcomes and acquisitions over the last five years.
They struck a bingo if they acquire 49% in chatgpt.
Moreover I think ultimately the GitHub play was to get more devs in the azure ecosystem too by providing first class integration between dev frontend (VSC), dev backend (WSL 2, devcontainers), corporate communication and calculus (Teams and Office365), GitHub, azure.
The only thing they are really missing out still is a good chat and videocall application.
Teams has good ideas and decent integrations, but it's goodness dies once you have to interact with the core of the application: video calls are mediocre and chat is plain terrible.
To me this is a sign they are no longer able to compete and are failing back, the only growth they expect is from acquisitions
This is not looking good
Betting on the right things is a hugely valuable skill.
See Android, Youtube, Instagram, Tesla ... All projects that someone believed in, bought them, put money in, and then they became huge.
All of the examples grew by 100x after being bought. Which of the Microsoft's acquisition grew by even 10x? Which of them even achieved breakeven in term of money spent?
How does it not matter? It's like convincing yourself that since you can order delivery food you're a chef. Being able to produce new ideas that appeal to people is the sought after skill Microsoft is literally paying for with money from their entrenched positions that are failing to make anything new and appealing. Buying things is a skill all companies have and Microsoft is not skilled for spending money to buy ideas they can't come up with
It is called "Opportunism"
"Look they are doing great things lately with AI", when they just pumped money into an already successful AI project
That's a short term vision, a sign of lack of capabilities in a specific domain, "let's buy our way in, nobody gonna notice"