Google debuts more powerful “Ultra 1.0” AI model in rebranded “Gemini” chatbot

Confusing name shuffles aside, "Gemini Advanced" vies to catch up with ChatGPT-4.

See full article...

 

[stormcrash]

It's not just products anymore, Google is even killing product names. On the other hand though Bard was always a terrible name, though Gemini says nothing about the product compared to Copilot (and wonder how often it will get misspelled Gemeni or something like that)

 

[alexrdavies]

The privacy policy on Gemini (at least in the European Economic Area) is both admirably clear, and scary.

Google collects your Gemini Apps conversations, related product usage information, info about your location and your feedback. Google uses this data, consistent with our Privacy PolicyOpens in a new window, to provide, improve and develop Google products and services and machine learning technologies, including Google’s enterprise products such as Google Cloud.

To help with quality and improve our products (such as generative machine learning models that power Gemini Apps), human reviewers read, annotate, and process your Gemini Apps conversations.

Conversations that have been reviewed or annotated by human reviewers (and related data like your language, device type, location info or feedback) are not deleted when you delete your Gemini Apps activity because they are kept separately and are not connected to your Google Account. Instead, they are retained for up to three years.

Even when Gemini Apps activity is off, your conversations will be saved with your account for up to 72 hours.

This is par for the course for Google - in general if Google and Microsoft provide the same service, Google keeps a lot more of the data - but worth repeating. This might be a good tool to play around with, but it isn't a place to put sensitive info.

 

[internetomancer]

Too bad. Strongly prefer the name "bard".

Short, simple, descriptive, and does not sound like a 1950's CIA robot supervillain.

 

[OkSureFine]

Odds on how long the name Gemini lasts?

 

[HiroTheProtagonist]

It's not just products anymore, Google is even killing product names. On the other hand though Bard was always a terrible name, though Gemini says nothing about the product compared to Copilot (and wonder how often it will get misspelled Gemeni or something like that)

I can sort of see it. "Bard" was a weird name, given the connotations of storytelling and song singing, but "Gemini" usually denotes duality and replication, which is pretty much all AI is good for.

 

[hambone]

The privacy policy on Gemini (at least in the European Economic Area) is both admirably clear, and scary.



This is par for the course for Google - in general if Google and Microsoft provide the same service, Google keeps a lot more of the data - but worth repeating. This might be a good tool to play around with, but it isn't a place to put sensitive info.

Talk about a no-privacy policy. Google employees reviewing an keeping Gemini chat transcripts is wild.

I can imagine warrants and subpoenas for Gemini chat transcripts are gonna become the new hotness in criminal and civil litigation.

And if a Google employee reviews a Gemini chat transcript that suggests someone is planning a murder or learning how to generate child porn, do they report it? Is there a duty to do so?

 

[citizencoyote]

I can sort of see it. "Bard" was a weird name, given the connotations of storytelling and song singing, but "Gemini" usually denotes duality and replication, which is pretty much all AI is good for.

Bards were also known for embellished and imaginative storytelling, which is something AI already struggles with (via hallucinations). I'm not surprised they moved away from the name.

 

[Emon]

Bards were also known for embellished and imaginative storytelling, which is something AI already struggles with (via hallucinations). I'm not surprised they moved away from the name.

It also avoids confusion about its capabilities, as machines are not yet able to perform magical music

 

[deltaproximus]

Ultra One (point zero)... Name is a little too close to Ultron, imo.

 

[jbriano]

Too bad. Strongly prefer the name "bard".

It gives you some idea what it does and doesn't sound like a 1950's CIA robot supervillain.

To each their own. I so thoroughly detested the name "Bard," I referred to it as "Google's Chat GPT-like AI offering"

 

[myownlittlworld]

Oh for Pete's sake, they couldn't just go and not steal another name from an existing project, again?

Gemini: https://gemini.circumlunar.space/

GoLang: https://en.wikipedia.org/wiki/Go!_(programming_language)

Why even bother namesquatting when it's only going to be around for 9 months?

 

[digraph]

Given the author's "huge grain of salt," why would anyone with even a shred of caution use a commercial LLM?

 

[internetomancer]

Given the author's "huge grain of salt," why would anyone with even a shred of caution use a commercial LLM?

Any and all cases where you don't need it to be "right".

So anything fun or trivial, or that can't be broken.

So, worst case:
"Write a legal brief."
Bad case:
"Write a C++ program that does..."
Okayish case:
"Why isn't my code working?"
Good case:
"Do you see any issues (spelling, grammar, flow, etc.) with this email?
Best case:
Let's play pretend...

 

[subscriber]

The privacy policy on Gemini (at least in the European Economic Area) is both admirably clear, and scary.



This is par for the course for Google - in general if Google and Microsoft provide the same service, Google keeps a lot more of the data - but worth repeating. This might be a good tool to play around with, but it isn't a place to put sensitive info.

Wow, I wish the article covered how crazy this privacy policy is.

 

[captain+obvious]

Odds on how long the name Gemini lasts?

I can see Google releasing messaging-app-2024 as "Gemini".

Can't wait to see Ron reporting it.

 

[AngryEssay]

What a confusing mess. Par for the course for Google.

 

[Species 8472]

They should call it Forbin

 

[AngryEssay]

To

To Google’s defence, I don’t think they kill products for the fun of it.

Google releases far more quirky and experimental products than any of its competitors . (they’re moving a bit away from that culture ) They tend to kill more underperforming products than competitors as well.

If you’re a tech company with limited (expensive) engineering resource, will you keep them working a product you know is ever going to make or move them something else ?

The alternative to is not releasing any new product at all.

They cancel products because their management is composed of tech bros on beanbags and Macbooks drinking pumpkin spice latte and throwing out phrases like "OKR" and "KPI" while not knowing what they actually mean. Talk about culture.

 

[xrmb]

Why Google, why? And I'm not ranting about the name change and different levels. But why introduce another level of Google One that is so different from the current plan. There is a 2TB plan, which gives 10% off in store and can be billed yearly to save. But now there is another 2TB plan that has neither of these features... I'd be willing to pay a little more to get the advanced AI, but I don't want to lose any of my existing benefits.

 

[stormcrash]

They cancel products because their management is composed of tech bros on beanbags and Macbooks drinking pumpkin spice latte and throwing out phrases like "OKR" and "KPI" while not knowing what they actually mean. Talk about culture.

Google's problem isn't bad engineering, it's bad program management that doesn't iterate on a product. But sure, blame your imaginary caricature of the spoiled hipster techbro instead

 

[zoward]

They should call it Forbin

I see what you did there...

 

[stjns]

To pay for Gemini Advanced, you have to sign up for a subscription plan called Google One

To clarify, the "Gemini Advanced" model is not included in current Google One subscriptions. I am a current Google One subscriber at the 2TB level, and I don't have access. In order to get access, I would need to switch from my current $99/year plan (which has an equivalent monthly plan at $9/mo) to a 2TB $19/mo plan, so this is effectively a $10/mo add on to the current 2TB plan.

 

[phat_tony]

To

To Google’s defence, I don’t think they kill products for the fun of it.

Google releases far more quirky and experimental products than any of its competitors . (they’re moving a bit away from that culture ) They tend to kill more underperforming products than competitors as well.

If you’re a tech company with limited (expensive) engineering resource, will you keep them working a product you know is ever going to make or move them something else ?

The alternative to is not releasing any new product at all.

Sorry for the cliché, but you must be new here. (/s. Yes, I saw your account goes back 10 years.)

Google isn't famous for killing products because they try things out and kill things that underperform.

They launch things all the time with apparently 0 thought for how the product fits into the business, or how they'd monetize it, and then they kill it even if it's very successful because they have no idea why they even did it to begin with. See Reader, Cloud Print, Inbox, Surveys, Picasa, etc. Then there's also all the products they kill or at least sabotage due to internal infighting or bad/incoherent corporate strategy - Hangouts, about 60 other messaging clients, Nest things, Google Pay, etc. Stadia they sabotaged internally vs their own initial plan, then hastily released to inevitable failure.

They are not like other companies. Anybody can launch something that fails and then kill it. Google kills even the things that succeed.

 

[thehardcard]

Odds on how long the name Gemini lasts?

Please! Google Gemini is a name your grandchildren will see. At least the ones born before 2005.

 

[Toastr]

To

To Google’s defence, I don’t think they kill products for the fun of it.

Google releases far more quirky and experimental products than any of its competitors . (they’re moving a bit away from that culture ) They tend to kill more underperforming products than competitors as well.

If you’re a tech company with limited (expensive) engineering resource, will you keep them working a product you know is ever going to make or move them something else ?

The alternative to is not releasing any new product at all.

Nah, that's a false dichotomy. I don't want Google to never cancel a product ever, I just want them to have a clear and coherent plan for supporting core products.
For example, I don't care if Google wants to experiment with a Zoom competitor, and I don't care if they try some new and bold things there and kill the project if things don't work out. But I don't want them to replace their official Android video call platform (twice!), especially if the new product is missing basic functionality present in the original.

 

[thehardcard]

What a confusing mess. Par for the course for Google.

Not just Google. Proper use of any commercial LLM requires you to ask “What version is this?”

 

[AngryEssay]

I can see Google releasing messaging-app-2024 as "Gemini".

Can't wait to see Ron reporting it.

I'm already predicting it:

"Gemini has been cancelled and will reach end-of-life on March 2nd. For Google Workspace users, you can transition to Google Chat"

"We're sorry to inform you that Google Chat is no longer supported. Click here to move your messages to Google Text Sending™"

"Google has announced that Text Sending has been shut down, in favor of Google Messages®"

 

[stormcrash]

Sorry for the cliché, but you must be new here. (/s. Yes, I saw your account goes back 10 years.)

Google isn't famous for killing products because they try things out and kill things that underperform.

They launch things all the time with apparently 0 thought for how the product fits into the business, or how they'd monetize it, and then they kill it even if it's very successful because they have no idea why they even did it to begin with. See Reader, Cloud Print, Inbox, Surveys, Picasa, etc. Then there's also all the products they kill or at least sabotage due to internal infighting or bad/incoherent corporate strategy - Hangouts, about 60 other messaging clients, Nest things, Google Pay, etc. Stadia they sabotaged internally vs their own initial plan, then hastily released to inevitable failure.

They are not like other companies. Anybody can launch something that fails and then kill it. Google kills even the things that succeed.

More importantly, rather than stick with something and fix its flaws they instead throw it out and let it rot in favor of a new shiny that doesn't do everything the old thing did and then has its own flaws. Each iteration they lose users who decide not to make the transition

 

[CEGalyon]

 

[kinpin]

Sorry for the cliché, but you must be new here. (/s. Yes, I saw your account goes back 10 years.)

Google isn't famous for killing products because they try things out and kill things that underperform.

They launch things all the time with apparently 0 thought for how the product fits into the business, or how they'd monetize it, and then they kill it even if it's very successful because they have no idea why they even did it to begin with. See Reader, Cloud Print, Inbox, Surveys, Picasa, etc. Then there's also all the products they kill or at least sabotage due to internal infighting or bad/incoherent corporate strategy - Hangouts, about 60 other messaging clients, Nest things, Google Pay, etc. Stadia they sabotaged internally vs their own initial plan, then hastily released to inevitable failure.

They are not like other companies. Anybody can launch something that fails and then kill it. Google kills even the things that succeed.

Care to give me some names of successful products they killed ?. Fact that you may find something useful doesn’t mean it’s hitting their targets.

Google Reader was killed more than 10 years ago, why isn’t anyone releasing a viable alternative? The answer is simple, it’s not a profitable product.

Then again there are times they buy companies (Eg. Picasa, Snapseed etc) purely for talent.

I can agree with you on Hangouts and their very disjointed messaging ecosystem .

 

[kinpin]

Nah, that's a false dichotomy. I don't want Google to never cancel a product ever, I just want them to have a clear and coherent plan for supporting core products.
For example, I don't care if Google wants to experiment with a Zoom competitor, and I don't care if they try some new and bold things there and kill the project if things don't work out. But I don't want them to replace their official Android video call platform (twice!), especially if the new product is missing basic functionality present in the original.

You have a point about their Video/ Messaging platforms lack of coherency. I think that’s one of the areas they’ve had terrible leadership.

 

[autoteleology]

Not confusing enough, they should have rebranded it as Android.

 

[Fatesrider]

Too bad. Strongly prefer the name "bard".

Short, simple, descriptive, and does not sound like a 1950's CIA robot supervillain.

I'm not fond of either name. I'd rather they call it HAL. They don't need the 9000 part.

Why?

Because neither humans, nor AI, can be trusted.

If the AI isn't fooling humans with an output, it's humans fooling humans with AI output. Unless AI can be made perfect, and not do things that fool humans such as saying, "I'm sorry, Dave, I can't do that" when someone provides the prompts in "nude Taylor Swift giving oral sex to Donald Trump photographic with studio lighting", then it's unreliable as anything other than a curiosity.

Skynet may be the generally un-seen villain of the future, but HAL is relatable on an individual level and a cultural cautionary touchstone for sci-fi buffs. People SHOULD be wary of what they produce. And people need to be more aware that they already produce a lot of stuff we see today.

HAL is the perfect name for these things. Call this one "HAL-Google". Call the others HAL-OpenAI, HAL-Microsoft, HAL-Apple, etc.

After all, AI's have produced one HAL of a mess already, and are poised to create even more trouble in the future.

 

[peterford]

What was formerly called "Duet AI" will become "Gemini for Google Workspace and Google Cloud."

That's awesome. Have they employed some old Microsoft staff nearing retirement?

 

[AngryEssay]

Google's problem isn't bad engineering, it's bad program management that doesn't iterate on a product. But sure, blame your imaginary caricature of the spoiled hipster techbro instead

Given the historical record, nothing about this is imaginary. It's just easier to think that random crypto ventures suffer from this issue more often, but the reality is that Google is as big of a mess as any other company. A corporation of that scale blundering that much is enough said.

 

[dcdp]

Why Google, why? And I'm not ranting about the name change and different levels. But why introduce another level of Google One that is so different from the current plan. There is a 2TB plan, which gives 10% off in store and can be billed yearly to save. But now there is another 2TB plan that has neither of these features... I'd be willing to pay a little more to get the advanced AI, but I don't want to lose any of my existing benefits.

The Premium AI plan builds on the existing 2TB plan (called the Premium Plan) by adding access to Gemini Advanced for an extra $10 a month (pricing is obviously different in different countries). However, it is only available month by month - having said that it is offering the first 2 months free (for me at least)
https://one.google.com/about/plans

 

[AngryEssay]

Sorry for the cliché, but you must be new here. (/s. Yes, I saw your account goes back 10 years.)

Google isn't famous for killing products because they try things out and kill things that underperform.

They launch things all the time with apparently 0 thought for how the product fits into the business, or how they'd monetize it, and then they kill it even if it's very successful because they have no idea why they even did it to begin with. See Reader, Cloud Print, Inbox, Surveys, Picasa, etc. Then there's also all the products they kill or at least sabotage due to internal infighting or bad/incoherent corporate strategy - Hangouts, about 60 other messaging clients, Nest things, Google Pay, etc. Stadia they sabotaged internally vs their own initial plan, then hastily released to inevitable failure.

They are not like other companies. Anybody can launch something that fails and then kill it. Google kills even the things that succeed.

I think it's interesting how some people here believe Google is supposed to be at the position and scale where things like that should not be happening, which is why they are so taken aback by the fact that it is just business as usual in valley bro land. They can simultaneously be a very large enterprise and also one that produces equally large miscalculations. The fact that this methodology has created Googles and Apples doesn't take away from the matter that it equally made happen FTXes and Theranos-es.