Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
481–490 of 1001 posts
Re: Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
#482I must say, I’m surprised to see so many HN users (who, despite being biased towards having more disposable income are also supposedly more discerning) simply immediately upgrading their Google One plans to the new offering before testing this and seeing how it fares. Looks like a potential gold mine for Google regardless of how it performs!
Re: Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
#483This is an impressive product, well done Google. There is a PM in there somewhere who knows what they are doing, kudos to you. Prediction: they get to 6-7 digit number of paying customers, decide it is peanuts for them (~$20M/mo) and instead decide to push the free version with ads with full force as the future of search.
Google announced they surpassed 100 million subscriptions to Google one already and $15B in revenue for subscriptions (between YouTube premium, TV and Google one). I'm not sure your estimate is realistic.
Re: Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
#484Did I get this right? Bard – old name of their generative AI service, to be called Gemini Duet AI – old name for their generative AI in Google Workspace, to be called Gemini Gemini – three things: 1. the name of their models (like GPT). 2. the new name of their free service (like ChatGPT), gives access to Pro 1.0 but not Ultra 1.0. 3. the new name of the Generative AI tools in Google Workspace. Gemini Advanced – the…
It’s sad that a company of very smart people can’t figure out coherent naming. Can you imagine Apple causing confusion like this? I know it’s not a like-for-like comparison, but everything Apple does it seems like they have a grand strategy that’s clear for everyone to see. Things build up in a modular way to fit a big puzzle. Google, on the other hand, constantly makes up things on an ad hoc basis.
Likewise for engineering, Google is organized into Product Areas (Geo, Search, Cloud, etc.), which also explains why one product would get some feature that would really make sense integrated into another... but it never happens.
Google is exceptionally good at making its products be near-perfectly reflective of its internal organization scheme. So reflective you can brush your teeth with it.
I'm often a broken record about this on HN - but IMO the PA organizational structure is a strong inhibitor to Google's success and ability to create coherent suites of products.
Re: Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
#485Did I get this right? Bard – old name of their generative AI service, to be called Gemini Duet AI – old name for their generative AI in Google Workspace, to be called Gemini Gemini – three things: 1. the name of their models (like GPT). 2. the new name of their free service (like ChatGPT), gives access to Pro 1.0 but not Ultra 1.0. 3. the new name of the Generative AI tools in Google Workspace. Gemini Advanced – the…
bard(RIP) was powered by gemini which was powered by palm2 or something, which was powered by deepmind which was powered by google which was powered by alphabet or sth Let s hope the name sticks for more than 2 months
2017: Transformers invented
2018: BERT
2020: Meena
2021: LaMDa -> First model Bard was built on
2022: PaLM
2023: PaLM2
Late 2023: Gemini
It probably would have been clearer if they used simple numerical versioning like OpenAI's GPT-{2,3,3.5,3}. I suppose the idea is to do that with Gemini now.More info here: https://en.wikipedia.org/wiki/Gemini_(chatbot)
Re: Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
#486Did I get this right? Bard – old name of their generative AI service, to be called Gemini Duet AI – old name for their generative AI in Google Workspace, to be called Gemini Gemini – three things: 1. the name of their models (like GPT). 2. the new name of their free service (like ChatGPT), gives access to Pro 1.0 but not Ultra 1.0. 3. the new name of the Generative AI tools in Google Workspace. Gemini Advanced – the…
Gemini is four things: The new mobile app is also called Google Gemini. Except on iOS, where Gemini is integrated into the Goople app instead. Also, while Gemini Advanced is supported in , this list is not the same as the the Google Gemini app is supported in on the Play Store. Make sure you check this before you spend money on your upgraded Google One subscription.
Re: Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
#487Earlier quoted context omitted.
It’s sad that a company of very smart people can’t figure out coherent naming. Can you imagine Apple causing confusion like this? I know it’s not a like-for-like comparison, but everything Apple does it seems like they have a grand strategy that’s clear for everyone to see. Things build up in a modular way to fit a big puzzle. Google, on the other hand, constantly makes up things on an ad hoc basis.
The naming on iPhones and Macbooks is terrible, and confusing now that the M3 Pro offers quite different CPU and bandwidth limits compared to M2 Pro...
The iPhone is literally just iPhone with Pro for high end, or no modifier for low end. Add “max” for big screen. The only confusion maybe is “max” isn’t obviously referring to screen size.
iPhone 15, 15 Pro, 15 max, iPhone 15 Pro Max.
Macs are the same way. I don’t think it’s fair to say it’s confusing that “M3” processor has different specs than “M2” processor.
Beyond that, Mac laptops are Pro vs Air, defining how powerful vs portable they are with associated screen size variant 14, 16 and 13,15.
Re: Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
#488Did I get this right? Bard – old name of their generative AI service, to be called Gemini Duet AI – old name for their generative AI in Google Workspace, to be called Gemini Gemini – three things: 1. the name of their models (like GPT). 2. the new name of their free service (like ChatGPT), gives access to Pro 1.0 but not Ultra 1.0. 3. the new name of the Generative AI tools in Google Workspace. Gemini Advanced – the…
I'm a Googler, and I think this is not only the right move but I've of the better names and renamings that the company has done. First, Gemini is just a better sounding name than Bard. And then, few users are going to care about the difference between the model and the app that lets you use the model. If Google kept this distinction they would have inevitably had to either come up with a new name for a new model, whi…
Cache invalidation and naming things are two of the hardest things in Computer Science, and so on.
By the way, there is still a Gemeni in there (“Bard becomes the *Gemeni* chat interface…”).
Re: Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
#489Earlier quoted context omitted.
It’s sad that a company of very smart people can’t figure out coherent naming. Can you imagine Apple causing confusion like this? I know it’s not a like-for-like comparison, but everything Apple does it seems like they have a grand strategy that’s clear for everyone to see. Things build up in a modular way to fit a big puzzle. Google, on the other hand, constantly makes up things on an ad hoc basis.
naming things is hard is one of the cultural difference between a computer-first logical mindset contrasted with the mathematics-first mindset the mathematics first mentality does not properly recognize the importance and the difficulty of having good names for things, whereas the computological view recognizes both: the importance and the difficulty
Re: Bard is now Gemini, and we’re rolling out a mobile app and Gemini Advanced
#490Earlier quoted context omitted.
If we want to test these beasts in logic, we should probably start using actual formalized logic, rather than English. In just one test, Gemini flopped hard, while GPT-4-Turbo nailed it. Here is my prompt: Below is a well-typed CoC function: foo : ∀(P: Nat -> *) ∀(s: ∀{n} -> ∀(x: (P n)) -> (P (n + 1))) ∀(z: (P 0)) (P 3) = λP λs λz (s (s (s z))) Below is an incomplete CoC function: foo : ∀(P: Nat -> *) ∀(f: ∀{n} -> ∀(…
ERRATA: I just noticed GPT-4 mixed up `h` and `f`, so it also got it wrong. This is a draw. Both AIs (and, apparently, myself) are terrible at reasoning. Guess we're not curing cancer with computers anytime soon :')