Gemini, the company’s family of generative AI models, will enhance the smart TV operating system so it can generate descriptions for movies and TV shows.
© 2024 TechCrunch. All rights reserved. For personal use only.
Google’s trying to make waves with Gemini, its flagship suite of generative AI models, apps and services.
So what is Gemini? How can you use it? And how does it stack up to the competition?
To make it easier to keep up with the latest Gemini developments, we’ve put together this handy guide, which we’ll keep updated as new Gemini models, features and news about Google’s plans for Gemini are released.
Gemini is Google’s long-promised, next-gen GenAI model family, developed by Google’s AI research labs DeepMind and Google Research. It comes in three flavors:
All Gemini models were trained to be “natively multimodal” — in other words, able to work with and use more than just words. They were pretrained and fine-tuned on a variety of audio, images and videos, a large set of codebases and text in different languages.
This sets Gemini apart from models such as Google’s own LaMDA, which was trained exclusively on text data. LaMDA can’t understand or generate anything other than text (e.g., essays, email drafts), but that isn’t the case with Gemini models.
Image Credits: Google
Google, proving once again that it lacks a knack for branding, didn’t make it clear from the outset that Gemini is separate and distinct from the Gemini apps on the web and mobile (formerly Bard). The Gemini apps are simply an interface through which certain Gemini models can be accessed — think of it as a client for Google’s GenAI.
Incidentally, the Gemini apps and models are also totally independent from Imagen 2, Google’s text-to-image model that’s available in some of the company’s dev tools and environments.
Because the Gemini models are multimodal, they can in theory perform a range of multimodal tasks, from transcribing speech to captioning images and videos to generating artwork. Some of these capabilities have reached the product stage yet (more on that later), and Google’s promising all of them — and more — at some point in the not-too-distant future.
Of course, it’s a bit hard to take the company at its word.
Google seriously underdelivered with the original Bard launch. And more recently it ruffled feathers with a video purporting to show Gemini’s capabilities that turned out to have been heavily doctored and was more or less aspirational.
Still, assuming Google is being more or less truthful with its claims, here’s what the different tiers of Gemini will be able to do once they reach their full potential:
Google says that Gemini Ultra — thanks to its multimodality — can be used to help with things like physics homework, solving problems step-by-step on a worksheet and pointing out possible mistakes in already filled-in answers.
Gemini Ultra can also be applied to tasks such as identifying scientific papers relevant to a particular problem, Google says — extracting information from those papers and “updating” a chart from one by generating the formulas necessary to re-create the chart with more recent data.
Gemini Ultra technically supports image generation, as alluded to earlier. But that capability hasn’t made its way into the productized version of the model yet — perhaps because the mechanism is more complex than how apps such as ChatGPT generate images. Rather than feed prompts to an image generator (like DALL-E 3, in ChatGPT’s case), Gemini outputs images “natively,” without an intermediary step.
Gemini Ultra is available as an API through Vertex AI, Google’s fully managed AI developer platform, and AI Studio, Google’s web-based tool for app and platform developers. It also powers the Gemini apps — but not for free. Access to Gemini Ultra through what Google calls Gemini Advanced requires subscribing to the Google One AI Premium Plan, priced at $20 per month.
The AI Premium Plan also connects Gemini to your wider Google Workspace account — think emails in Gmail, documents in Docs, presentations in Sheets and Google Meet recordings. That’s useful for, say, summarizing emails or having Gemini capture notes during a video call.
Google says that Gemini Pro is an improvement over LaMDA in its reasoning, planning and understanding capabilities.
An independent study by Carnegie Mellon and BerriAI researchers found that the initial version of Gemini Pro was indeed better than OpenAI’s GPT-3.5 at handling longer and more complex reasoning chains. But the study also found that, like all large language models, this version of Gemini Pro particularly struggled with mathematics problems involving several digits, and users found examples of bad reasoning and obvious mistakes.
Google promised remedies, though — and the first arrived in the form of Gemini 1.5 Pro.
Designed to be a drop-in replacement, Gemini 1.5 Pro is improved in a number of areas compared with its predecessor, perhaps most significantly in the amount of data that it can process. Gemini 1.5 Pro can take in ~700,000 words, or ~30,000 lines of code — 35x the amount Gemini 1.0 Pro can handle. And — the model being multimodal — it’s not limited to text. Gemini 1.5 Pro can analyze up to 11 hours of audio or an hour of video in a variety of different languages, albeit slowly (e.g., searching for a scene in a one-hour video takes 30 seconds to a minute of processing).
Gemini 1.5 Pro entered public preview on Vertex AI in April.
An additional endpoint, Gemini Pro Vision, can process text and imagery — including photos and video — and output text along the lines of OpenAI’s GPT-4 with Vision model.
Using Gemini Pro in Vertex AI. Image Credits: Gemini
Within Vertex AI, developers can customize Gemini Pro to specific contexts and use cases using a fine-tuning or “grounding” process. Gemini Pro can also be connected to external, third-party APIs to perform particular actions.
In AI Studio, there’s workflows for creating structured chat prompts using Gemini Pro. Developers have access to both Gemini Pro and the Gemini Pro Vision endpoints, and they can adjust the model temperature to control the output’s creative range and provide examples to give tone and style instructions — and also tune the safety settings.
Gemini Nano is a much smaller version of the Gemini Pro and Ultra models, and it’s efficient enough to run directly on (some) phones instead of sending the task to a server somewhere. So far, it powers a couple of features on the Pixel 8 Pro, Pixel 8 and Samsung Galaxy S24, including Summarize in Recorder and Smart Reply in Gboard.
The Recorder app, which lets users push a button to record and transcribe audio, includes a Gemini-powered summary of your recorded conversations, interviews, presentations and other snippets. Users get these summaries even if they don’t have a signal or Wi-Fi connection available — and in a nod to privacy, no data leaves their phone in the process.
Gemini Nano is also in Gboard, Google’s keyboard app. There, it powers a feature called Smart Reply, which helps to suggest the next thing you’ll want to say when having a conversation in a messaging app. The feature initially only works with WhatsApp but will come to more apps over time, Google says.
And in the Google Messages app on supported devices, Nano enables Magic Compose, which can craft messages in styles like “excited,” “formal” and “lyrical.”
Google has several times touted Gemini’s superiority on benchmarks, claiming that Gemini Ultra exceeds current state-of-the-art results on “30 of the 32 widely used academic benchmarks used in large language model research and development.” The company says that Gemini 1.5 Pro, meanwhile, is more capable at tasks like summarizing content, brainstorming and writing than Gemini Ultra in some scenarios; presumably this will change with the release of the next Ultra model.
But leaving aside the question of whether benchmarks really indicate a better model, the scores Google points to appear to be only marginally better than OpenAI’s corresponding models. And — as mentioned earlier — some early impressions haven’t been great, with users and academics pointing out that the older version of Gemini Pro tends to get basic facts wrong, struggles with translations and gives poor coding suggestions.
Gemini 1.5 Pro is free to use in the Gemini apps and, for now, AI Studio and Vertex AI.
Once Gemini 1.5 Pro exits preview in Vertex, however, the model will cost $0.0025 per character while output will cost $0.00005 per character. Vertex customers pay per 1,000 characters (about 140 to 250 words) and, in the case of models like Gemini Pro Vision, per image ($0.0025).
Let’s assume a 500-word article contains 2,000 characters. Summarizing that article with Gemini 1.5 Pro would cost $5. Meanwhile, generating an article of a similar length would cost $0.1.
Ultra pricing has yet to be announced.
The easiest place to experience Gemini Pro is in the Gemini apps. Pro and Ultra are answering queries in a range of languages.
Gemini Pro and Ultra are also accessible in preview in Vertex AI via an API. The API is free to use “within limits” for the time being and supports certain regions, including Europe, as well as features like chat functionality and filtering.
Elsewhere, Gemini Pro and Ultra can be found in AI Studio. Using the service, developers can iterate prompts and Gemini-based chatbots and then get API keys to use them in their apps — or export the code to a more fully featured IDE.
Code Assist (formerly Duet AI for Developers), Google’s suite of AI-powered assistance tools for code completion and generation, is using Gemini models. Developers can perform “large-scale” changes across codebases, for example updating cross-file dependencies and reviewing large chunks of code.
Google’s brought Gemini models to its dev tools for Chrome and Firebase mobile dev platform, and its database creation and management tools. And it’s launched new security products underpinned by Gemini, like Gemini in Threat Intelligence, a component of Google’s Mandiant cybersecurity platform that can analyze large portions of potentially malicious code and let users perform natural language searches for ongoing threats or indicators of compromise.
Gemini Nano is on the Pixel 8 Pro, Pixel 8 and Samsung Galaxy S24 — and will come to other devices in the future. Developers interested in incorporating the model into their Android apps can sign up for a sneak peek.
It might! Apple and Google are reportedly in talks to put Gemini to use for a number of features to be included in an upcoming iOS update later this year. Nothing’s definitive, as Apple is also reportedly in talks with OpenAI, and has been working on developing its own GenAI capabilities.
This post was originally published Feb. 16, 2024 and has since been updated to include new information about Gemini and Google’s plans for it.
Gemini 1.5 Pro, Google’s most capable generative AI model, is now available in public preview on Vertex AI, Google’s enterprise-focused AI development platform. The company announced the news during its annual Cloud Next conference, which is taking place in Las Vegas this week.
Gemini 1.5 Pro launched in February, joining Google’s Gemini family of generative AI models. Undoubtedly its headlining feature is the amount of context that it can process: between 128,000 tokens to up to 1 million tokens, where “tokens” refers to subdivided bits of raw data (like the syllables “fan,” “tas” and “tic” in the word “fantastic”).
One million tokens is equivalent to around 700,000 words or around 30,000 lines of code. It’s about four times the amount of data that Anthropic’s flagship model, Claude 3, can take as input and about eight times as high as OpenAI’s GPT-4 Turbo max context.
A model’s context, or context window, refers to the initial set of data (e.g. text) the model considers before generating output (e.g. additional text). A simple question — “Who won the 2020 U.S. presidential election?” — can serve as context, as can a movie script, email, essay or e-book.
Models with small context windows tend to “forget” the content of even very recent conversations, leading them to veer off topic. This isn’t necessarily so with models with large contexts. And, as an added upside, large-context models can better grasp the narrative flow of data they take in, generate contextually richer responses and reduce the need for fine-tuning and factual grounding — hypothetically, at least.
So what specifically can one do with a 1 million-token context window? Lots of things, Google promises, like analyzing a code library, “reasoning across” lengthy documents and holding long conversations with a chatbot.
Because Gemini 1.5 Pro is multilingual — and multimodal in the sense that it’s able to understand images and videos and, as of Tuesday, audio streams in addition to text — the model can also analyze and compare content in media like TV shows, movies, radio broadcasts, conference call recordings and more across different languages. One million tokens translates to about an hour of video or around 11 hours of audio.
Thanks to its audio-processing capabilities, Gemini 1.5 Pro can generate transcriptions for video clips, as well, although the jury’s out on the quality of those transcriptions.
In a prerecorded demo earlier this year, Google showed Gemini 1.5 Pro searching the transcript of the Apollo 11 moon landing telecast (which comes to about 400 pages) for quotes containing jokes, and then finding a scene in movie footage that looked similar to a pencil sketch.
Google says that early users of Gemini 1.5 Pro — including United Wholesale Mortgage, TBS and Replit — are leveraging the large context window for tasks spanning mortgage underwriting; automating metadata tagging on media archives; and generating, explaining and transforming code.
Gemini 1.5 Pro doesn’t process a million tokens at the snap of a finger. In the aforementioned demos, each search took between 20 seconds and a minute to complete — far longer than the average ChatGPT query.
Google previously said that latency is an area of focus, though, and that it’s working to “optimize” Gemini 1.5 Pro as time goes on.
Of note, Gemini 1.5 Pro is slowly making its way to other parts of Google’s corporate product ecosystem, with the company announcing Tuesday that the model (in private preview) will power new features in Code Assist, Google’s generative AI coding assistance tool. Developers can now perform “large-scale” changes across codebases, Google says, for example updating cross-file dependencies and reviewing large chunks of code.
Alphabet CEO Sundar Pichai took the stage on Wednesday at a Stanford event held by the university’s business school, offering some small insights into what he thinks about running one of the world’s most valuable tech companies.
It was a notable appearance because Pichai’s been having a bit of a rough go lately. Google is widely perceived to have gotten a late start on generative AI, trailing behind Microsoft-funded OpenAI. That’s despite the fact that the company under Pichai has been focusing on AI for the better part of the last decade, and Google researchers wrote the formative paper on transformer models that really kicked off the generative AI revolution. More recently, Alphabet’s Gemini LLM was excoriated for generating bizarrely inaccurate images of historical situations, such as depicting America’s founding fathers as Black or Native American, rather than white English men, suggesting an overcorrection for certain types of bias.
The interviewer, Stanford Graduate School of Business dean Jonathan Levin, wasn’t exactly a hostile inquisitor — at the end, he revealed that the two men’s sons had once played in a middle school band together — and Pichai is deft at answering difficult questions by posing them as further questions about how he thinks, rather than with direct answers. But there were a couple nuggets of interest during the talk.
At one point, Levin asked what Pichai tried to do to keep a company of 200,000 people innovating against all the startups battling to disrupt its business. It’s obviously something Pichai worries about.
“Honestly, it’s a question which has always kept me up at night through the years,” he started. “One of the inherent characteristics of technology is you can always develop something amazing with a small team from the outside. And history has shown that. Scale doesn’t always give you … Regulators may not agree, but at least running the company, I’ve always felt you’re always susceptible to someone in a garage with a better idea. So I think, how do you as a company move fast? How do you have the culture of risk-taking? How do you incent for that? These are all things which you actually have to work at a lot. I think at least larger organizations tend to default. One of the most counterintuitive things I’ve seen is, the more successful things are, the more risk averse people become. It’s so counterintuitive. You would often find smaller companies almost make decisions which bet the company, but the bigger you are, it’s true for large university, it’s true a large company, you have a lot more to lose, or you perceive you have a lot more to lose. And so you find you don’t take as many ambitious risk-taking initiatives. So you have to consciously do that. You have to push teams to do that.”
He didn’t offer any specific tactics that have proven successful at Google, but instead noted how difficult it is to create the proper incentives.
“One example for this is … how do you reward effort and risk-taking and good execution, and not always outcomes? It’s easy to think you should reward outcomes. But then people start gaming it, right? People take conservative things in which you will get a good outcome.”
He hearkened back to an earlier time in which Google was more willing to take weird risks, in particular pointing to the firm’s ill-fated Google Glass; it didn’t work out, but it was one of the first devices to experiment with augmented reality.
“We recently said, we went back to a notion we had in early Google of Google Labs. And so we’re setting a thing up by which it’s easier to put out something without always worrying about, you know, the full brand and the weight of building a Google product. How can you put out something in the easy way, the lighter weight way? How do you allow people to prototype more easily internally and get it out to people?”
Later, Levin asked what advances Pichai was most excited about this year.
First, he cited the multimodality of Google’s latest LLM — that is, its ability to process different kinds of inputs, such as video and text, simultaneously.
“All our AI models now already are using Gemini 1.5 Pro; that’s a 1 million context window and it’s multimodal. The ability to process huge amounts of information in any type of modality on the input side and give it on the output side, I think it’s mind blowing in a way that we haven’t fully processed.”
Second, he highlighted the ability to connect different discrete answers to provide smarter workflows. “Where today you’re using the LLMs as just an information-seeking thing, but chaining them together in a way that you can kind of tackle workflows, that’s going to be extraordinarily powerful. It could maybe make your billing system in Stanford Hospital a bit easier,” he joked.
You can watch the entire interview, along with an interview with Fed Chairman Jerome Powell that happened prior to it, on YouTube. Levin and Pichai start around 1 hour and 18 minutes in.
Google’s trying to make waves with Gemini, a new generative AI platform that recently made its big debut. But while Gemini appears to be promising in a few aspects, it’s falling short in others. So what is Gemini? How can you use it? And how does it stack up to the competition?
To make it easier to keep up with the latest Gemini developments, we’ve put together this handy guide, which we’ll keep updated as new Gemini models and features are released.
Gemini is Google’s long-promised, next-gen generative AI model family, developed by Google’s AI research labs DeepMind and Google Research. It comes in three flavors:
All Gemini models were trained to be “natively multimodal” — in other words, able to work with and use more than just text. They were pre-trained and fine-tuned on a variety audio, images and videos, a large set of codebases, and text in different languages.
That sets Gemini apart from models such as Google’s own large language model LaMDA, which was only trained on text data. LaMDA can’t understand or generate anything other than text (e.g. essays, email drafts and so on) — but that isn’t the case with Gemini models. Their ability to understand images, audio and other modalities is still limited, but it’s better than nothing.
Image Credits: Google
Google, proving once again that it lacks a knack for branding, didn’t make it clear from the outset that Gemini is separate and distinct from Bard. Bard is simply an interface through which certain Gemini models can be accessed — think of it as an app or client for Gemini and other gen AI models. Gemini, on the other hand, is a family of models — not an app or frontend. There’s no standalone Gemini experience, nor will there likely ever be. If you were to compare to OpenAI’s products, Bard corresponds to ChatGPT, OpenAI’s popular conversational AI app, and Gemini corresponds to the language model that powers it, which in ChatGPT’s case is GPT-3.5 or 4.
Incidentally, Gemini is also totally independent from Imagen-2, a text-to-image model that may or may not fit into the company’s overall AI strategy. Don’t worry, you’re not the only one confused by this!
Because the Gemini models are multimodal, they can in theory perform a range of tasks, from transcribing speech to captioning images and videos to generating artwork. Few of these capabilities have reached the product stage yet (more on that later), but Google’s promising all of them — and more — at some point in the not-too-distant future.
Of course, it’s a bit hard to take the company at its word.
Google seriously under-delivered with the original Bard launch. And more recently it ruffled feathers with a video purporting to show Gemini’s capabilities that turned out to have been heavily doctored and was more or less aspirational. Gemini is, to the tech giant’s credit, available in some form today — but a rather limited form.
Still, assuming Google is being more or less truthful with its claims, here’s what the different tiers of Gemini models will be able to do once they’re released:
Few people have gotten their hands on Gemini Ultra, the “foundation” model on which the others are built, so far — just a “select set” of customers across a handful of Google apps and services. That won’t change until sometime later this year, when Google’s largest model launches more broadly. Most info about Ultra has come from Google-led product demos, so it’s best taken with a grain of salt.
Google says that Gemini Ultra can be used to help with things like physics homework, solving problems step-by-step on a worksheet and pointing out possible mistakes in already filled-in answers. Gemini Ultra can also be applied to tasks such as identifying scientific papers relevant to a particular problem, Google says — extracting information from those papers and “updating” a chart from one by generating the formulas necessary to recreate the chart with more recent data.
Gemini Ultra technically supports image generation, as alluded to earlier. But that capability won’t make its way into the productized version of the model at launch, according to Google — perhaps because the mechanism is more complex than how apps such as ChatGPT generate images. Rather than feed prompts to an image generator (like DALL-E 3, in ChatGPT’s case), Gemini outputs images “natively” without an intermediary step.
Unlike Gemini Ultra, Gemini Pro is available publicly today. But confusingly, its capabilities depend on where it’s used.
Google says that in Bard, where Gemini Pro launched first in text-only form, the model is an improvement over LaMDA in its reasoning, planning and understanding capabilities. An independent study by Carnegie Mellon and BerriAI researchers found that Gemini Pro is indeed better than OpenAI’s GPT-3.5 at handling longer and more complex reasoning chains.
But the study also found that, like all large language models, Gemini Pro particularly struggles with math problems involving several digits, and users have found plenty of examples of bad reasoning and mistakes. It made plenty of factual errors for simple queries like who won the latest Oscars. Google has promised improvements, but it’s not clear when they’ll arrive.
Gemini Pro is also available via API in Vertex AI, Google’s fully managed AI developer platform, which accepts text as input and generates text as output. An additional endpoint, Gemini Pro Vision, can process text and imagery — including photos and video — and output text along the lines of OpenAI’s GPT-4 with Vision model.
Using Gemini Pro in Vertex AI.
Within Vertex AI, developers can customize Gemini Pro to specific contexts and use cases using a fine-tuning or “grounding” process. Gemini Pro can also be connected to external, third-party APIs to perform particular actions.
Sometime in “early 2024,” Vertex customers will be able to tap Gemini Pro to power custom-built conversational voice and chat agents (i.e. chatbots). Gemini Pro will also become an option for driving search summarization, recommendation and answer generation features in Vertex AI, drawing on documents across modalities (e.g. PDFs, images) from different sources (e.g. OneDrive, Salesforce) to satisfy queries.
Image Credits: Gemini
In AI Studio, Google’s web-based tool for app and platform developers, there’s workflows for creating freeform, structured and chat prompts using Gemini Pro. Developers have access to both Gemini Pro and the Gemini Pro Vision endpoints, and they can adjust the model temperature to control the output’s creative range and provide examples to give tone and style instructions — and also tune the safety settings.
Gemini Nano is a much smaller version of the Gemini Pro and Ultra models, and it’s efficient enough to run directly on (some) phones instead of sending the task to a server somewhere. So far it powers two features on the Pixel 8 Pro: Summarize in Recorder and Smart Reply in Gboard.
The Recorder app, which lets users push a button to record and transcribe audio, includes a Gemini-powered summary of your recorded conversations, interviews, presentations and other snippets. Users get these summaries even if they don’t have a signal or Wi-Fi connection available — and in a nod to privacy, no data leaves their phone in the process.
Gemini Nano is also in Gboard, Google’s keyboard app, as a developer preview. There, it powers a feature called Smart Reply, which helps to suggest the next thing you’ll want to say when having a conversation in a messaging app. The feature initially only works with WhatsApp, but will come to more apps in 2024, Google says.
There’s no way to know how the Gemini family really stacks up until Google releases Ultra later this year, but the company has claimed improvements on the state of the art — which is usually OpenAI’s GPT-4.
Google has several times touted Gemini’s superiority on benchmarks, claiming that Gemini Ultra exceeds current state-of-the-art results on “30 of the 32 widely used academic benchmarks used in large language model research and development.” The company says that Gemini Pro, meanwhile, is more capable at tasks like summarizing content, brainstorming and writing than GPT-3.5.
But leaving aside the question of whether benchmarks really indicate a better model, the scores Google points to appear to be only marginally better than OpenAI’s corresponding models. And — as mentioned earlier — some early impressions haven’t been great, with users and academics pointing out that Gemini Pro tends to get basic facts wrong, struggles with translations, and gives poor coding suggestions.
Gemini Pro is free to use in Bard and, for now, AI Studio and Vertex AI.
Once Gemini Pro exits preview in Vertex, however, the model will cost $0.0025 per character while output will cost $0.00005 per character. Vertex customers pay per 1,000 characters (about 140 to 250 words) and, in the case of models like Gemini Pro Vision, per image ($0.0025).
Let’s assume a 500-word article contains 2,000 characters. Summarizing that article with Gemini Pro would cost $5. Meanwhile, generating an article of a similar length would cost $0.1.
The easiest place to experience Gemini Pro is in Bard. A fine-tuned version of Pro is answering text-based Bard queries in English in the U.S. right now, with additional languages and supported countries set to arrive down the line.
Gemini Pro is also accessible in preview in Vertex AI via an API. The API is free to use “within limits” for the time being and supports 38 languages and regions including Europe, as well as features like chat functionality and filtering.
Elsewhere, Gemini Pro can be found in AI Studio. Using the service, developers can iterate prompts and Gemini-based chatbots and then get API keys to use them in their apps — or export the code to a more fully featured IDE.
Duet AI for Developers, Google’s suite of AI-powered assistance tools for code completion and generation, will start using a Gemini model in the coming weeks. And Google plans to bring Gemini models to dev tools for Chrome and its Firebase mobile dev platform around the same time, in early 2024.
Gemini Nano is on the Pixel 8 Pro — and will come to other devices in the future. Developers interested in incorporating the model into their Android apps can sign up for a sneak peek.
We’ll keep this post up to date with the latest developments.
Founded in 2017, Atoms Lanka Solutions is a provider of IT consulting and software development services. Throughout the company’s history, we have worked with over 160 customers from 5+ countries..
