Amidst the mountains of vendor cheerleading for generative AI efforts, often amplified by enterprise board members, skeptical CIOs tend to feel outnumbered. But their cynical worries may now have some company, in the form of a report from Apple and an interview from Meta — both of which raise serious questions about whether genAI can actually do much of what its backers claim.
The debate involves some fairly amorphous terms, at least when spoken in a computing environment context — things like reasoning and logic. When a large language model (LLM), for example, proposes a different and ostensibly better way to do something, is it because its sophisticated algorithm has figured out a better way? Or is it just wildly guessing, and sometimes it gets lucky? Or did it hallucinate something and accidentally say something helpful?
Would a CIO ever trust a human employee with such tendencies? Not likely, but IT leaders are regularly tasked with integrating genAI tools into the enterprise environment by corporate executives expecting miracles.
The conclusions drawn by AI experts from Apple and Meta may help CIOs set more realistic expectations about what genAI models can and cannot do, now and in the near future.
“Our findings reveal that LLMs exhibit noticeable variance when responding to different instantiations of the same question. Specifically, the performance of all models declines when only the numerical values in the question are altered.
“Furthermore, we investigate the fragility of mathematical reasoning in these models and demonstrate that their performance significantly deteriorates as the number of clauses in a question increases… When we add a single clause that appears relevant to the question, we observe significant performance drops (up to 65%) across all state-of-the-art models, even though the added clause does not contribute to the reasoning chain needed to reach the final answer.”
What does mathematical reasoning have to do with AI-powered business applications? The Apple research team spelled it out:
“Mathematical reasoning is a crucial cognitive skill that supports problem-solving in numerous scientific and practical applications. Consequently, the ability of large language models (LLMs) to effectively perform mathematical reasoning tasks is key to advancing artificial intelligence and its real-world applications.”
What today’s state-of-the-art LLMs do is not logical reasoning, the researchers concluded:
“Current LLMs are not capable of genuine logical reasoning; instead, they attempt to replicate the reasoning steps observed in their training data… It may resemble sophisticated pattern matching more than true logical reasoning.”
“When a departing OpenAI researcher in May talked up the need to learn how to control ultra-intelligent AI, LeCun pounced. ‘It seems to me that before “urgently figuring out how to control AI systems much smarter than us,” we need to have the beginning of a hint of a design for a system smarter than a house cat,’ he replied on X.
“He likes the cat metaphor. Felines, after all, have a mental model of the physical world, persistent memory, some reasoning ability and a capacity for planning, he says. None of these qualities are present in today’s ‘frontier’ AIs, including those made by Meta itself.”
Later, the WSJ story lets LeCun make his central point:
“Today’s models are really just predicting the next word in a text, he says. But they’re so good at this that they fool us. And because of their enormous memory capacity, they can seem to be reasoning, when in fact they’re merely regurgitating information they’ve already been trained on.
“‘We are used to the idea that people or entities that can express themselves, or manipulate language, are smart — but that’s not true,’ says LeCun. ‘You can manipulate language and not be smart, and that’s basically what LLMs are demonstrating.’”
That is the key issue. Enterprises are putting far too much faith in genAI systems, says Francesco Perticarari, general partner at technology investment house Silicon Roundabout Ventures in London, England.
It’s easy to assume that the rare correct answers given by these tools are flashes of brilliance, rather than the genAI having gotten a lucky guess. But “the output is not based at all on reasoning. It is merely based on extremely powerful computing,” Perticarari said.
Putting genAI in the driver’s seat
One frequently cited selling point for genAI is that some models have proven quite effective at passing various state bar exams. But those bar exams are ideal environments for genAI, because the answers are all published. Memorizations and regurgitation are ideal uses for genAI, but that doesn’t mean genAI tools have the skills, understanding, and intuition to practice law.
“The logic is that if genAI can pass the bar exam, it can handle my business, build systems that are robust and that work now,” said Alan Nichol, co-founder and CTO of AI vendor Rasa. “[Business leaders] are taking this dangerous, naive approach and just letting the LLM figure it out,” he said.
Nichol pointed to Apple’s analysis that the more complex and multilayered math problems got, the more the LLMs got lost and confused.
“It’s supposed to understand this math, but something is definitely fishy. The medium through which they are doing [these calculations] is natural language. It’s fuzzy and imprecise,” he said. “Language models were never supposed to do a lot of these things. There are vanishingly few situations where you want your software to guess what it should be doing, what the next few steps should be.”
Nichol stressed that these systems, left to their own devices, are reckless. “Four out of five times, genAI doesn’t follow its own instructions,” he said. “You want it to guess business logic? It just doesn’t work and is extremely slow and consumes a tremendous amount of tokens.”
Perticarari from Silicon Roundabout Ventures is especially concerned about hallucinations coupled with the lack of meaningful guardrails. GenAI seems to easily overcome — or be tricked by a user into overcoming — many of the safeguards organizations attempt to place around it.
“If you have a one-year-old, you wouldn’t give her a loaded gun and then try and explain to her why she shouldn’t shoot you,” Perticarari said. “[GenAI is] not sentient. Humans are sentient and they assume the system is intelligent, too. Letting genAI run on autopilot to me is crazy. Don’t give anything to a black box.”
Fighting FOMO
Perticarari blames enterprise executives and board members for falling victim to countless AI sales pitches. He says that CIOs have to be the voice of sanity.
“It is always easy during a gold rush to sell hype. [Sales execs] just keep delivering endless layers of selling without really understanding,” Perticarari said. “CIOs need to ask, ‘How fundamental and vital is the task that [we] are outsourcing to genAI?’”
Jake Reynolds, the CTO at cybersecurity vendor Wirespeed, agrees. He maintains that a lot of the rush to genAI has been pushed by board members, and “the CIO had to tag along.”
Executives are giving in to FOMO (fear of missing out), thinking that “their largest competitor is doing it, so we are going to do it,” he said. “But it doesn’t deliver. Even with the more objective mathematics, it starts falling apart. Try to get consistency out of it. You can’t. The words it predicts changes every time you tweak a little knob… Are you really OK with your product only working 80% of the time?”
Reynolds encourages CIOs to slow down and be as minimalistic as practical. “We’re not laggards. We’re just realists about what the technology can really do,” he said.
Judicious use of genAI tools can mitigate disappointment or worse, agrees Nichol. “We should just let the LLMs do what the LLMs are amazing at. Don’t let the LLM do everything.”
The US government has announced new rules restricting investments in China’s AI and other tech sectors deemed threats to national security, expanding the existing restrictions that were so far limited to exports.
First introduced by the US Treasury in June, the rules are based on an executive order signed by President Joe Biden in August 2023.
They focus on three critical areas: semiconductors and microelectronics, quantum information technologies, and certain AI systems.
“This narrow set of technologies is core to the next generation of military, cybersecurity, surveillance, and intelligence applications,” the Treasury said in a statement.
The US already restricts or bans the export of many technologies covered by the new rules to certain countries. The new program complements existing export controls and inbound screening measures by blocking US investments from aiding the development of sensitive technologies in countries of concern, the Treasury added.
Fueling the trade war
This marks the latest development in the ongoing trade war between the US and China, which has already witnessed numerous restrictions.
Analysts are skeptical of the policy’s impact, cautioning that it may further intensify tensions and stifle innovation and growth.
“The scope of restrictions is now expanding beyond the sale of technology IP or chips to include investments in the Chinese tech sector,” said Neil Shah, VP of research and partner at Counterpoint Research. “This move aims to stifle Chinese tech companies on both fronts — limiting financial and technology inflows. Unfortunately, this will make it difficult for Chinese companies to innovate quickly and will further intensify the geopolitical tech cold war.”
This also means that if China retaliates — while protecting its own manufacturing ecosystem — it could affect large and small tech companies that still rely on China as a key market.
Restrictions could stifle collaboration and knowledge exchange between nations, potentially slowing innovation by reducing opportunities to work on advanced projects.
“Companies might also need to reassess their strategic priorities, which may lead to an unnecessary increase in innovation costs,” said Charlie Dai, VP and principal analyst at Forrester. “On the other hand, regulatory concerns will force enterprises outside the US to further prioritize localization strategies to achieve self-sufficiency in critical areas, potentially leading to increasingly isolated innovation ecosystems.”
The new rules may also require US enterprises to closely monitor both domestic and international regulatory shifts and establish agile compliance programs to adapt swiftly to evolving requirements.
“These constraints can also diminish R&D investments and have profound long-term economic effects, stifling advancement in pivotal sectors like semiconductors, quantum computing, and AI, ultimately hampering overall technological progress,” said Thomas George, president of Cybermedia Research.
Opportunity for emerging markets
For other emerging markets, however, the tightened US restrictions could present new opportunities by attracting redirected foreign investments from US firms.
“As trade tensions rise and new regulations emerge, US companies increasingly move away from Chinese manufacturing,” said George. “Instead, they want to collaborate with countries such as India, Mexico, and Vietnam. This shift is crucial as it enhances companies’ resilience and allows them to navigate new US export controls more effectively.”
Companies should reduce dependency on any single country by diversifying supply chains to mitigate risks associated with regulatory changes in specific regions, according to Dai.
“Engaging with research and advisory firms can help them better understand the potential impact of various regulatory changes, prepare contingency plans, and develop strategies to assess and mitigate risks,” Dai said.
The Phoenix fabrication facility of the world’s largest semiconductor chip maker is yielding more usable chips than similar plants in Taiwan, according to the Taipei Times.
Rick Cassidy, president of TSMC’s US division, said during a webinar last week that the share of usable chips from the company’s Phoenix plant exceeds that of similar Taiwanese plants by 4%. If true, the superior performance at the Phoenix fab is notable because the US government has been working to spur a return of the semiconductor manufacturing industry to US shores, where manufacturing tends to be more costly.
Better yields would help offset those higher costs.
“Four percent higher yield is certainly good news,” said Harry Moser, president of the Reshoring Initiative, a non-profit that offers companies assessments on offshoring costs. “To be competitive, we need a higher yield. It is agreed that US factory capital cost and operating cost will be 10% to 20% higher than in most other countries. The 4% will offset some of that difference.”
The COVID-19 pandemic highlighted critical gaps in the semiconductor supply chain as imports to the US and other nations ground to a halt, affecting the production of everything with electronics, from smart phones to cars. The CHIPS and Science Act, passed in 2022, earmarked more than $52 billion in funding and tax incentives for use by the US semiconductor industry to create new or expand existing manufacturing and R&D facilities.
The CHIPS Act was created to address both future possible supply chain catastrophes and to re-establish the US as a major chips manufacturer.
To date, the CHIPS Act has allocated more than $32 billion in proposed funding across 18 companies, 16 states, and 26 projects. However, no CHIPS funding has yet been disbursed to any companies, according to the US Department of Commerce.
TSMC is the main supplier of chips for both Nvidia and Apple. The CHIPS Act allocated $6.6 billion in grants and $5 billion in loans, along with a 25% tax credit, to incentivize the company to build three fabs in Arizona. TSMC’s first facility was scheduled to open this year, but the company pushed that back to next year after labor shortages surfaced.
The US reshoring efforts come at a time when the industry doesn’t have anywhere near the workforce — including technicians, computer scientists, and engineers — required to support future needs. By some estimates, the US semiconductor industry will face a worker shortfall of between 59,000 and 146,000 workers by 2029. A minimum of 50,000 trained semiconductor engineers will be needed over the next several years in the US to meet the overwhelming and rapidly growing demand, according to a study by Purdue University.
The broader US economy is set to have a gap of 1.4 million such workers, according to a 2023 study from the Semiconductor Industry Association. So the competition will be fierce over those skilled workers. Compounding the problem is an ongoing exodus of existing talent as older workers retire. A study from Deloitte found that nearly 90% of tech leaders interviewed cited recruiting as their biggest challenge.
A TSMC spokesperson shared statements regarding the Phoenix fab with Computerworld from a third quarter earnings call by CEO C.C. Wei, but declined to comment on Cassidy’s claim directly.
“Our first fab entered engineering wafer production in April with 4-nanometer process technology, and the result is a highly satisfactory, with a very good yield,” Wei said. “This is an important operational milestone for TSMC and our customers, demonstrating TSMC’s strong manufacturing capability and execution.”
Wei said he expects volume production of the company’s first Arizona fab to start in early 2025, and he is “confident” it will deliver “the same level of manufacturing quality and reliability” from our fabs in Taiwan.
TSMC is also building two other fabs in the Phoenix area that will use more advanced technologies based on its customer needs, Wei said. The second fab is scheduled to begin volume production in 2028 and the third fab will begin production by the end of the decade.
“Thus, TSMC will continue to play a critical and integral role in enabling our customers’ success, while remaining a key partner and enabler of the US semiconductor industry,” Wei said.
Reshore Now’s Moser said it would be good to know whether the Phoenix fab uses identical equipment as in Taiwan, speculating that the US plant could have been more modern. “Was it accomplished solely by US workers or significantly by Taiwanese brought over to aid the start-up?” he said.
Apple’s week of Mac began today with a newly announced iMac, now beefed up with an M4 chip and more internal memory. Apple says the iMac is up to 4.5 times faster than an equivalent all-in-one Intel Core 7 Windows PC — and promises the machine will deliver up to six times the performance of the most popular Intel-based iMac.
The inference is obvious: if you use your iMac professionally, you might want to think about an upgrade. Reinforcing the point, Apple says the iMac is up to 1.7 times faster for most tasks and 2.1 times faster for more advanced tasks when compared to the M1 model.
Apple sets the scene for its AI
Part of the reason for this improved performance is the big boost to 16GB of unified onboard memory (configurable to 24GB). That memory boost is to support Apple Intelligence, which is also available for Macs running macOS Sequoia 15.1 or above. The Neural Engine in M4 chips is 3x faster than the M1. (Apple Intelligence is also now available for iPhones and iPads running iOS/iPad OS 18.1.)
Be warned, the entry-level $1,299 iMac might not reach these performance heights as it ships with an 8-core CPU; the rest of the range offers 10 cores. You do get hardware-accelerated ray tracing, which is going to make a big difference when you use the Mac.
Apple’s AI platform play means all its current devices will now support Apple Intelligence, meaning the company now offers the world’s biggest AI ecosystem.
What else is new?
The iMac display continues to be the same 24-in. 4.5K Retina display we all know and love, with a new nano-texture glass option available if color fidelity and anti-reflection matters to you.
Starting at $1,299, the new iMacs are available in a “parade” of colors, including green, yellow, orange, pink, purple, blue, and silver. Buyers get: a 12-megapixel Center Stage camera with support for Desk View; a brilliant microphone and speaker system; and four USB-C ports, all of which support Thunderbolt 4. You can even run two external displays. Wrapping it up, you’ll find Wi-Fi 6E and Bluetooth 5.1 along with TouchID support, thanks to a button on the keyboard.
Overall, this is a solid and update, but it has to be said that it doesn’t seem to be the main attraction — that honor this week is likely to be all about a smaller machine….
We’re still waiting for the Mac mini
I can’t help but feel the iMac is being seen through a lens of pre-announcement speculation for the Mac mini. That product is already attracting lots of interest — just look at the pre-release headlines:
“This is the Mac Mini’s big moment” (The Verge).
“A tiny Mac mini could be the ultimate travel companion and I can’t wait for it” (TechRadar).
“Apple Mac mini with M4 chip could be a game-changer for creatives, here’s why” (Hindustan Times).
Talk about setting the scene.
Even Bloomberg’s Mark Gurman has put his well-connected assessment out there; he obliquely tells us that even if you are quite happy with your M1 Mac mini, the move to M4 processors “could feel as significant as that first shift from Intel machines to Apple chips.”
That’s borne out by Apple’s iMac claims above. What we can piece together from the iMac introduction is that the new Mac mini will also deliver huge performance boosts in contrast to the M1 or M2 models already in use.
That’s an upgrade productivity benefits are built on in some industries, and it suggests that if your business has M1 (or older) Mac minis in its fleet, the new M4 models seem to be a tempting upgrade. After all, you don’t even need to replace the display….
Will a M4 Mac mini be the new Mac for business?
This introduction is expected to be about more than the silicon inside these Macs — it’s also the new design around them. If reports are correct, the new mini may be significantly smaller as Apple’s designers draw yet another benefit out of the energy and heat dissipation advantages of the company’s Arm-based chips.
Expect it to be a small aluminium box that’s taller but otherwise similar in size to the current Apple TV. I visualize this as being a box about half the size of a regular paperback book and perhaps as thick as three average length novels stacked atop each other. That’s really small. And it should now come with 16GB of base memory and support for Apple Intelligence.
Speaking just last year, MacStadium CTO Chris Chapman told me his existing server farms full of Mac minis used so much less power that his data center providers were, “always calling us up to tell us we’re not using enough power for the space.”
If the smaller size means lower energy consumption (and given what we know of Apple’s silicon evolution so far, it probably does), then for enterprises handling hundreds of these machines — or any other Mac, come to that — the M4 upgrade promises significant reductions in energy costs.
A good start to a week of Mac
Combined with the faster chip, these tiny desktop Mac minis or larger iMacs are going to run just about anything you want as effectively as a hot knife through butter.
That’s why the upcoming Mac mini has generated so much interest, even before its introduction. Combined with the impressive iMac rollout today and anticipation around the expected powerful MacBook Pro improvements, Apple’s big week of Mac news is off to a strong start. But will it distract or focus interest on the company’s end of year results announcement Thursday?
OpenAI plans to release its next major flagship model, Orion, by December, according to The Verge — and Microsoft, which has invested heavily in Open AI, is said to be ready to launch Orion on Azure as early as November.
Orion is intended to be the successor to GPT-4 and is said to be up to 100 times stronger. However, unlike Open AI’s last two models, GPT-4o and o1, it will not be released first through the AI assistant Chat GPT. Instead, OpenAI plans to first give a collection of companies access to the AI model so they can use it to build their own products.
It’s unclear whether the new model will be called GPT-5 or something else. Both Open AI and Microsoft declined to comment on the report, though OpenAI spokesperson Niko Felix told The Verge the company doesn’t “have plans to release a model code-named Orion this year.”
Microsoft has moved to simplify how Office app files are opened on mobile devices after users cited uncertainty around the process.
Opening a Word, Excel, or PowerPoint file on a mobile device typically means the Microsoft 365 mobile app will start up rather than the relevant standalone app.
This has been the case since Microsoft introduced an Office mobile app for Android and iOS in 2019. The Office app combined Word, Excel, and PowerPoint apps into a single app to create an “integrated experience” and reduce the need to switch between tools, Microsoft said at the time.
After user feedback, Microsoft announced last week that the standalone apps will be used to open files instead — as long as the standalone Office apps are installed on a user’s device; if not, files will continue to open in the Microsoft 365 app.
Aside from reducing confusion, opening the standalone app will make it easier to multitask, Microsoft said in a blog post, with multiple Office apps open at once.
“For customers who want to open more than one Word, Excel, or PowerPoint file at once, the standalone apps can better handle side by side and windowing scenarios that modern tablet and mobile operating systems support,” said Samer Sawaya, a principal product manager at Microsoft.
Microsoft has started apply the change across its Office apps for iOS and Android, beginning with OneDrive. Outlook will follow this month and in November, while the timing for Teams has yet to be confirmed.
Workspace is Google’s suite of productivity software tools, the main competitor to market leader Microsoft 365 (formerly Office 365). Formerly known as G Suite, Google Workspace includes well-known apps such as Gmail, Docs, Sheets, and Slides — apps that are also used for free by billions of people globally.
Google has been selling productivity apps to businesses for almost two decades now, having helped pave the way for cloud-based office productivity apps in the 2000s. At a time when most office suites were based on installed desktop software and sold as one-time purchases, Google bundled its popular free web apps with various business services to create a cloud-based office software subscription — a move that would eventually push competitors like Microsoft to adopt a subscription model and develop web versions of their apps.
Google was “very much ahead of the game” at this point and “even 11 or 12 years ago when Office 365 lagged behind what Google was doing with SaaS-based enterprise services,” said Joe Mariano, director analyst and member of Gartner’s Employee Experience Technology team.
The flexibility of Workspace’s cloud-hosted apps was then and continues to be a key draw for business customers, said Mariano: “The proposition is going to a completely SaaS-based set of services,” which allows for “deep interconnectivity” between the various components of the Workspace suite. “Because everything is based in the cloud on the Google Workspace architecture, you see a lot more seamless integration between those services,” he said.
These days, Google’s top priority for Workspace is the integration of generative AI features via its Gemini AI assistant (formerly Duet AI), introduced last year. The company has long been a major player in AI research and began to integrate AI features into its apps well before the recent generative AI boom — smart replies being one early example. With Gemini now available to embed into a range of Workspace apps (for an additional monthly fee), Google promises to save office workers time on tasks like writing emails, summarizing documents, and tracking down information.
In this article:
What is Google Workspace?
What happened to G Suite?
What are the different Google Workspace plans?
What add-ons are available for Google Workspace?
What is Gemini for Google Workspace?
How does Google Workspace adoption compare with Microsoft 365?
How easy is it to move from Microsoft 365 to Google Workspace?
What are some of the newest Google Workspace features?
What’s next for Google Workspace?
What is Google Workspace?
Workspace is Google’s suite of cloud-based productivity and collaboration apps, typically sold as a paid monthly subscription to business and enterprise customers.
Apps included in most Workspace plans cover core business productivity and collaboration use cases: email (Gmail), document editing (Docs), spreadsheets (Sheets), presentations (Slides), videoconferencing (Meet), team messaging (Chat), and scheduling (Calendar), to name just a handful of the available apps. Paid add-ons are available too, such as the Gemini AI assistant, Google Voice telephony services, and premium features of the no-code development platform AppSheet.
Workspace apps are accessed via a browser or mobile app, with no native desktop apps, aside from document storage app Google Drive. It’s possible to save and access files without an internet connection, though, once offline access is enabled for Drive. Users can also edit Docs, Sheets, and Slides files offline in the Chrome browser.
Although Google offers many of these same apps to consumers for free, paid Workspace subscriptions include additional features and business services such as a custom email domain, higher storage limits, and shared team storage. Google also provides a range of security, management, and analytics tools for IT admins in most paid Workspace subscriptions — with more advanced tools available on pricier plans aimed at large enterprise organizations. These range from device management to data retention tools, employee usage analytics, and more.
Not all Workspace plans are paid, however: Google in 2021 expanded the brand to bring some features that were formerly reserved for paid plans, such as smart canvases, to the free consumer versions of its apps under the banner of “Google Workspace for everyone.” Google also offers a limited free version of Workspace for business customers, called Google Workspace Essentials Starter.
The list of available apps in Workspace is somewhat shorter than that of Microsoft 365. “Google Workspace doesn’t reach feature parity with M365; it offers perhaps 60% to 70% of the features,” said J.P. Gownder, vice president and principal analyst on Forrester’s Future of Work team, though he describes the generative AI tools in Gemini for Google Workspace as “robust.”
Providing a more streamlined set of applications can be a good thing for business users, said Gartner’s Mariano. “Google doesn’t build new products, they build new services into their existing products, so it creates a lot less confusion,” he said. Take Gmail, for example. “You can live in your Gmail window all day if you want to: you can start writing new documents from there, you can check your calendar, you can start Chat, you can check Google Drive. The amount of digital friction that cuts back on is very, very significant.”
What happened to G Suite?
Although many business users still think of Google’s app suite as “G Suite,” the old name has gone — the Workspace brand replaced G Suite in 2020.
The switch to Workspace was not the first name change for Google’s bundle of work apps, which has roots back in the mid ’00s. Google Apps for Your Domain was the first iteration in 2006: a free ad-supported suite that included Gmail, Google Calendar, Page Creator (later Sites), and the Google Talk instant messaging app.
In 2007, the company began offering a paid ad-free subscription service aimed at businesses; initially called Google Apps Premier, it was later renamed Google Apps for Business and then Google Apps for Work. In 2016, the company announced its new G Suite brand with great fanfare, only to replace it four years later with the Workspace rebranding.
What are the different Google Workspace plans?
There are numerous payment plans for Google Workspace, targeted at everything from the smallest of small businesses up to large enterprises with hundreds of thousands of employees.
Google Workspace Individual: Aimed at sole traders/entrepreneurs, the Individual plan costs $8.33 per month on an annual basis and provides access to apps such as Gmail, Drive, Calendar, and Meet. The plan includes 1TB of storage and features such as appointment booking, live streaming on YouTube, and custom email layouts, as well as access to support. Custom email addresses are not included, so your address will end in “@gmail.com.”
Business Starter, Business Standard, and Business Plus: Targeted at small and midsize businesses, the three plans have a limit of 300 users and cost from $6 per user a month up to $18 per user a month. The lowest tier, Starter, provides 30GB of storage per user, custom email addresses, and Meet video calls with up to 100 participants.
All three Business plans include access to Workspace apps, but certain features — such as Meet breakout rooms — are available only on the more expensive tiers. Each Business plan includes endpoint management features, but advanced tools, such as Google’s Vault information governance and e-discovery, require Standard and Plus subscriptions.
Enterprise Standard and Plus: Suited to large businesses, there’s no cap on the number of users for the two Enterprise plans. Each user is assigned 5TB of storage, though more can be requested, while the Meet meeting cap is lifted to 1,000 participants.
Enterprise plans offer a wider range of security and management tools than Business tiers, including data loss prevention and advanced security analytics via the Workspace security center. The Enterprise Plus plan is the top Workspace tier, with features such as client-side encryption and advanced data exports. Pricing information isn’t publicly available.
Essentials Starter, Enterprise Essentials, and Enterprise Essentials Plus: These Workspace plans lower subscription costs (or drop them entirely) by removing Gmail. Essentials Starter is a free service aimed at business customers that provides up to 100 users with access to core Workspace apps. There’s little in the way of management features, and there are usage restrictions such as three Meet video call participants at a time.
Enterprise Essentials and Enterprise Essentials Plus — aimed at larger organizations — include premium app features, access to security and management tools, and no limit on the number of users. The Essentials Plus plan offers more storage per user and a wider range of advanced management tools and app features. Google doesn’t publish pricing for its Enterprise Essentials or Essentials Plus plans.
Note: A previous plan, called simply “Essentials,” is no longer available to new customers.
Frontline Starter and Frontline Standard: There are two Workspace editions for staff in roles such as customer service, retail, and manufacturing who do most of their work away from a desk. Both include core Workspace apps and management tools available in the Business editions. The Standard plan includes a wider range of security and admin tools, such as data loss prevention and Google Vault eDiscovery. Google doesn’t publish pricing for its Workspace Frontline plans.
Workspace Education Fundaments is free for qualifying institutions, while Education Standard costs $3 per student per year and includes additional security and admin features. Teaching and Learning Upgrade costs $4 per staff member per month for Education Fundamentals and Standard customers; it provides access to extra Classroom and video-communication features. The Education Plus plan is $5 per student per year and includes all the aforementioned features as well as organizational branding and more.
Workspace for Nonprofits, Workspace for Government, Workspace for various verticals: Google offers discounted pricing of its paid products to qualifying nonprofit and government organizations. Workspace for Government can be configured to meet FedRAMP High compliance standards required by US government organizations. Google also offers industry-specific editions of its Enterprise plans tailored for healthcare and life sciences, retail, manufacturing, professional services, and technology firms. Contact Google for pricing.
What add-ons are available for Google Workspace?
There are numerous add-ons for Workspace that, in most cases, require an extra monthly fee on top of core subscription costs. These add-ons provide access to some of the latest Workspace features and can broaden the suite’s functionality significantly, with the potential to increase a customer’s monthly spending at the same time.
Access to Gemini, Google’s generative AI assistant, adds an extra $30 per user a month to Workspace Enterprise subscriptions, for example. Google also provides a business telephony service, Voice, that costs from $10 per user per month for the Starter version up to $30 per user a month for the Premium version.
No-code development platform AppSheet Core is included in most paid Workspace subscriptions, but full functionality requires an additional fee. Similarly, Google Chrome Enterprise provides additional security and management tools over the free Core version for $6 per user per month.
Among the other add-ons is “Chat interoperability” — a free service that lets Workspace Chat users send messages to other collaboration tools such as Teams and Slack.
What is Gemini for Google Workspace?
Gemini is the name of Google’s generative AI assistant; it’s available as a standalone chatbot and can be embedded into various Workspace apps. It’s built on Google’s own Gemini AI models, such as Gemini 1.5 Pro and Imagen 3. The AI assistant was made generally available for business users in August 2023 under the name Duet AI, before switching to the Gemini branding in February this year.
Gemini for Google Workspace is an add-on subscription that gives business users access to premium Gemini features. There are two main versions of Gemini for Workspace: Enterprise and the less expensive Business tier, which launched in February this year.
Until recently, a Gemini for Google Workspace subscription was required for access to both the standalone Gemini app and the embedded version of the assistant. The Gemini app is now included with Workspace business plans, but the ability to interact with Gemini in the Workspace apps still requires a Gemini for Google Workspace subscription.
And in-app use is arguably where the assistant is most useful. Gemini can summarize a lengthy document in Docs, for example, draft an email in Gmail, create images in Slides with Google’s Imagen 3 text-to-video model, and plenty more besides. Google outlines numerous potential business use cases, from marketing professionals generating campaign briefs to HR workers creating job descriptions.
The business versions of Gemini offer many of the same features as the consumer versions of Gemini. One key difference is that Google won’t access data from user prompts to improve its own products and to train its AI models on the paid business versions, according to Google.
Gemini Business is available as a paid add-on across all Workspace subscriptions and costs $20 per user a month. For Gemini Business there’s a cap of 1,000 interactions — where Gemini is directed to summarize an email, or redraft a document, for example — with the AI assistant per user each month.
Gemini Enterprise tier, which costs $30 per user each month, is available as a paid add-on for all Workspace plans, apart from Business Starter and Individual. It includes access to a wider range of features, such as live translated captions and automated note taking in Meet video calls. It also removes the cap on user interactions each month.
Google has also made the AI assistant available to higher education institutions with Gemini Education ($20 per user each month) and Education Premium ($30 per user each month).
Not all Workspace customers require the full list of Gemini functionality for staff, and, for many workers, AI assistants have proved most effective in and around meetings. To cater to these customers, Google launched an add-on “AI Meetings and Messages” that includes access to Gemini features in Meet only, for lower price of $10 per user a month.
There’s also an “AI Security” add-on, also costing $10 per user per month, that provides access to AI-powered document classification in Google Drive. (The feature is also available in the more expensive Gemini Enterprise subscription.)
How does Google Workspace adoption compare with Microsoft 365?
Although Microsoft has retained its formidable lead in the market, Google has succeeded in establishing Workspace as a competitive office software suite. It’s hard to say exactly how Workspace adoption has fared in comparison with Microsoft 365 and other productivity app suites, but there are some indications.
Google reports 10 million paying customers for Workspace, while Microsoft claims more than 400 million paid commercial seats (end users) for Office 365/Microsoft 365. Google says its figure refers to the total number of organizations of any size that pay for Workspace, rather than the total number of paid seats/users across customer organizations, but this likely includes some extremely small — even one-person — businesses, which makes a comparison difficult.
The other stat that Google provides — of 3 billion users — is less useful, as it mostly relates to consumer use of its free apps. “While billions of consumers use free versions of Google Workspace applications, its use in enterprises remains limited — dwarfed by the popularity of Microsoft 365,” said Gownder.
Financial statements from Google’s parent company Alphabet do little to shed light on Workspace business adoption, as Workspace revenues are part of the “Google Cloud” segment that also includes Google Cloud Platform, Google’s wider cloud computing infrastructure and analytics service. In an earnings call for Alphabet’s most recent quarterly financial statement (Q2 FY24), CEO Sundar Pichai only noted that a 29% year-on-year increase in Google Cloud revenues was due, in part, to “strong Workspace growth.” This growth was mostly down to increases in average revenue per seat, said Pichai — an indication that, like others in the space, Google sees more value in increasing sales with existing customers than expanding its overall customer base.
Nevertheless, Google is gaining ground on M365, said Mariano, albeit slowly. “I do believe we will see continued market growth for Google Workspace,” he said. “Although it’s incremental, we do see that they are getting a little bit of market share every year.”
Google has managed to lure large enterprises such as Verizon and Airbus away from its rival, and, more recently, announced that 250,000 US Army personnel would be given Workspace licenses, a further sign of the suite’s acceptance by large organizations.
“When we look at the Western world, we see that Microsoft and Google are the two dominant players in this space,” Mariano said.
A key driver for business interest in Workspace is the cost compared to M365, said Mariano. Businesses are wary of the rising expenditure on Microsoft’s suite, he said, even if the reality is that, for the most part, businesses are unlikely to see significant savings by switching to Workspace.
“The term that I have heard from multiple IT leaders and CIOs I’ve talked to is the cost increase of Microsoft 365 is not sustainable,” said Mariano. “That is making them think much harder about Google Workspace than they have in the past.”
Gownder also sees Google’s efforts to target Workspace at educational institutions over the years starting to pay off. “Younger cohorts of employees, because they used Google Workspace in school, might have a preference for the suite over the less familiar Microsoft 365,” he said.
Another factor is that, after emerging from a reactive crisis mode during the pandemic, IT leaders are now in a better position to evaluate their workplace app strategy.
“When we look back at the pandemic era, Microsoft feasted: everybody ran to Microsoft, because it’s what they had,” Mariano said. “I think the dust has settled, and a lot of IT leaders and CIOs are looking at what’s left after the fact now and saying, ‘What did we do, and should we be here?’ They’re taking the time to decide what the next five to 10 years are going to look like.”
How easy is it to move from Microsoft 365 to Google Workspace?
For many businesses, adopting Workspace will mean migrating users from Microsoft’s business productivity apps. Google provides tools to help customers make the transition.
The Google Workspace Migrate tool lets admins move large amounts of content into a Workspace domain. Workspace Migrate can be used to move data from Microsoft Exchange (2010, 2013, 2016, and 2019), Exchange Online, Microsoft SharePoint (2010, 2013, and 2016), SharePoint Online, and Microsoft OneDrive for Business. It can also be used to migrate data from cloud storage platform Box, file share systems, and other Workspace domains. Workspace Migrate is available for Workspace Business Standard and Business Plus, Enterprise Standard and Enterprise Plus, and Education Standard and Education Plus.
For smaller migrations from certain Microsoft products, Google has other options: Google Workspace Migration for Microsoft Exchange (GWMME), Google Workspace Migration for Microsoft Outlook (GWMMO), and the data migration service.
Google also introduced a new file migration service in open beta recently that aims to simplify the transfer of files between from OneDrive to Google Drive for up to 100 employees at a time, and promises “minimal disruption” to end users during the process.
Nevertheless, a full migration from Microsoft 365 to Google Workspace is not for the faint-hearted. Such a project can require a “monumental effort” for a large enterprise, said Gownder: “There’s just a lot of complexity to manage, despite the assistive tools Google offers,” he said.
One of the challenges businesses face is that some aspects of the Microsoft 365 suite don’t have a clear destination in Google Workspace. For example, it’s hard to migrate Teams data in a way that makes sense, said Gownder, while Excel has entire programming sequences, macros, and other content approaches that can break when migrating them to Sheets. “Large organizations would want to work with an external consulting partner, and it won’t be a small project,” he said.
What are some of the newest Google Workspace features?
At the top of the agenda for Google is the integration of its Gemini AI assistant across Workspace apps.
One recently launched Gemini for Workspace feature in Meet is “Take notes for me,” which tasks the AI assistant with note taking during a video call, allowing participants to concentrate on their meeting.
Gemini can take notes in Google Meet meetings.
Google
A recent update to Gemini in Gmail’s “Help me write” feature — which uses generative AI to draft an email — adds three options to refine the text of an email: Formalize, Elaborate, and Shorten.
Gems are also now available for Gemini users. Similar to OpenAI’s GPT, Gems are customized versions of the AI chat assistant that can be instructed to respond in certain ways: coding partner and writing editor are examples of two pre-built Gems provided by Google.
In terms of the core Workspace apps, one fairly recent addition is “smart chips,” an array of interactive elements you can embed in Docs and Sheets to enhance collaboration. When a co-worker clicks on an embedded chip, they can either see more information, such as the contact details for a person or business, or take an action, such as changing a task’s status via a dropdown menu.
loading="lazy" width="400px">A people smart chip reveals contact info for that person.
Howard Wen / IDG
Currently under development is a new Workspace app: Vids. This is a video creation app that guides users through the process of producing simple videos for purposes such as sales pitches, employee onboarding, and learning and development. An AI assistant in the app can create a storyboard and suggest background images, for instance, and offers pre-set voiceovers to narrate a video. The app is now in preview in Workspace Labs.
What’s next for Google Workspace?
Although Gemini has been built into a wide range of Workspace apps already, it’s fair to say that Google will continue to focus its efforts on developing the AI assistant for work use.
“I think it’s Gemini, Gemini, Gemini for them right now,” said Mariano.
One AI feature in development is the AI Teammate for Workspace — an AI agent that human workers can interact directly with in the Workspace Chat app. The prototype, unveiled at Google I/O this year, could be used for a variety of purposes — to monitor and track projects, for example.
It remains to be seen whether Google’s investments in Gemini will result in attracting more businesses to the Workspace suite, however. “Google continues to bolster Workspace with AI-based tools, but we are not yet at the point where enterprises are saying they will migrate to Google Workspace because of Gemini,” said Gownder.
One strength for Google could lie in the combination of its Gemini and Workspace apps with its hardware business, with Google’s Pixel phones able to run AI on-device. “That is a competitive advantage for them that we could see grow in the next couple years, as we see what happens with AI,” said Mariano.
It’s not all about AI, though: Another intriguing product in the works is Project Starline, a 3D videoconferencing system that gives users the feeling that they’re in the same room. Google recently announced that it plans to make the system, which it has been developing for several years, commercially available in 2025.
Apple will introduce new Macs and the first services within its Apple Intelligence collection next week. To protect cloud-based requests made through Apple Intelligence, it has put industry-beating security and privacy protecting transparency in place around cloud-based requests handled by its Private Cloud Compute (PCC) system.
What that means is that Apple has pulled far ahead of the industry in a bid to build rock-solid protection around security and privacy for requests made of AI using Apple’s cloud. It’s an industry-leading move and is already delighting security researchers.
Why is that? It’s because Apple has opened the doors that secure its Private Cloud Compute systemwide to security testers in the hope that the energy of the entire infosec community will combine to help build a moat to protect the future of AI.
Make no mistake, this is what is at stake.
As AI promises to permeate everything, the choice we face is between a future of surveillance the likes of which we have never seen before, or the most powerful machine/human augmentation we can dream of. Server-based AI promises both these futures, even before mentioning that as quantum computing looms just a few hills and valleys away, the information picked up by non-private AI systems can be weaponized and exploited in ways we can’t even imagine.
That means to be secure tomorrow we must take steps today.
Protecting AI in the cloud
In part, protecting that future and ensuring it can say with total confidence that Apple Intelligence is the world’s most private and secure form of AI is what Apple is trying to do with PCC. This is the system that lets Apple run generative AI (genAI) models that need more processing power than available on the iPad, iPhone, or Mac you use to get things done. It’s the first port of call for these AI requests and has been deliberately designed to protect privacy and security. “You should not have to hand over all the details of your life to be warehoused and analyzed in someone’s AI cloud,” Apple Senior Vice President of Software Engineering Craig Federighi said when announcing PCC at WWDC.
The company promised that to “build public trust” in its cloud-based AI systems, it would allow security and privacy researchers to inspect and verify the end-to-end security and privacy of the system. The reason the security community is so excited is because Apple has exceeded that promise by making public all the resources it made available to researchers.
Security research for the rest of us
It provided the following resources:
The PCC Security Guide
Apple has published the PCC Security Guide, an extensive 100-page document including comprehensive technical details about the components of the system and how they work together to secure AI processing in the cloud. This is a deep guide that discusses built-in hardware protections and how the system handles various attack scenarios.
A Virtual Research Environment
The company has also created something security researchers might get excited about: A Virtual Research Environment (VRE) for the Apple platform. This consists of a set of tools that make it possible to perform your own security analysis of PCC using a Mac. This is a robust testing environment that runs a PCC node — basically a production machine — in a VM so you can beat it up as much as you like in search of security and privacy flaws.
You can use these tools to:
List and inspect PCC software releases.
Verify the consistency of the transparency log.
Download the binaries corresponding to each release.
Boot a release in a virtualized environment.
Perform inference against demonstration models.
Modify and debug the PCC software to enable deeper investigation.
Publishing the PCC source code
This is a big step in its own right and is provided under a license agreement that lets researchers dig deep for flaws. Within this set of information, the company has made source code that covers privacy, validation, and logging components. (All of this source code is available on GitHub now.)
Bounty hunters
Of course, the company understands that it must also incentivize researchers. To do so, Apple opened up a bounty system for those who succeed in finding flaws in the PCC code.
To contextualize the extent of this commitment, it is important to note that the value of these bounties is equal to what the company pays to researchers who discover iOS security flaws.
I believe that means Apple sees AI as a very important component to its future, PCC as an essential hub to drive forward to tomorrow, and that it will also now find some way to transform platform security using similar tools. Apple’s fearsome reputation for security means even its opponents have nothing but respect for the robust platforms it has made. That reputation is also why more and more enterprises are, or should be, moving to Apple’s platforms.
The mantle of protecting security is now under the passionate leadership of Ivan Krstić, who also led the design and implementation of key security tools such as Lockdown Mode, Advanced Data Protection for iCloud, and two-factor authentication for Apple ID. Krstić has previously promised that, “Apple runs one of the most sophisticated security engineering operations in the world, and we will continue to work tirelessly to protect our users from abusive state-sponsored actors like NSO Group.”
When it comes to bounties for uncovering flaws in PCC, researchers can now earn up to $1 million dollars if they find a weakness that allows arbitrary code execution with arbitrary entitlements, or a cool $250,000 if they uncover some way to access a user’s request data or sensitive information about their requests.
There are many other categories, and Apple seems really committed to ensuring it motivates even trivial discoveries: “Because we care deeply about any compromise to user privacy or security, we will consider any security issue that has a significant impact to PCC for an Apple Security Bounty reward, even if it doesn’t match a published category,” the company explains. Apple will award the biggest bounties for vulnerabilities that compromise user data and inference request data.
Apple’s gamble
It’s important to stress that in moving to deliver this degree of industry-leading transparency, Apple is gambling it can ensure that any weaknesses that do exist in its solution will be spotted and revealed, rather than being identified only to be sold on or weaponized.
The thinking is that while nation state-backed attackers might have access to resources that provide attackers with similar breadth of insight into Apple’s security protections, they will not share word of any such vulnerabilities with Apple. Such attackers, and those in the most well financed criminal or semi-criminal entities (within which I personally believe surveillance-as-a-service mercenaries belong), will spend time and money finding vulnerabilities in order to exploit them.
But there is a big world of security researchers who might also uncover weaknesses in the system who would be willing to share them, enabling Apple to patch vulnerabilities faster.
The way Apple sees it is that one way to ensure such vulnerabilities aren’t turned into privacy-destroying attacks is to make it so more people discover them at the same time; after all, even if one dodgy researcher chooses to use a weakness in an attack, another might disclose it to Apple early, effectively shutting down that route. In other words, by making these details available, Apple changes the game. In a strange irony, making these security protections open and available may well serve to make them more secure.
That’s the hope, anyway.
“We believe Private Cloud Compute is the most advanced security architecture ever deployed for cloud AI compute at scale, and we look forward to working with the research community to build trust in the system and make it even more secure and private over time,” Apple explained.
Why this matters
It is also a defining moment in security for AI. Why? Because Apple is an industry leader that sets expectations with its actions. With these actions, the company just defined the degree of transparency to which all companies offering cloud-based AI systems should now be held. If Apple can, they can, too. And any business or individual whose data or requests is being handled by cloud based AI systems can now legitimately demand that degree of transparency and protection. Apple is making waves again.
Humanoid robots from Tesla can “be a teacher, babysit your kids…, walk your dog, mow your lawn, get the groceries, just be your friend, serve drinks. Whatever you can think of, it will do,” said Elon Musk at this month’s Tesla “We, Robot” showcase in Burbank, CA.
To paraphrase: “Humanoid robots can be human, so you don’t have to!”
It’s a toxic and misguided vision for humanity — let me tell you why.
What is a robot, anyway?
When the public thinks of a robot, they think of a humanoid robot — an intelligent machine with feet, legs, knees, arms, a neck, a head, and a face — thanks to countless movies, TV shows, novels, and comics.
But what is a robot, really?
A robot is any computerized machine with sensors that gather information from its environment and can perform actions in the physical world based on that input. Robots’ attributes include autonomy, programmability, sensing capability, decision-making, and the ability to perform physical actions.
We talk about robots as a kind of future technology. But we’re surrounded by them every day.
The world of robots is big (so bear with me here); it includes consumer camera drones, smart thermostats, automated garage door openers, self-checkout kiosks, modern washing machines and dryers, robotic lawnmowers, automated pool cleaners, smart refrigerators, advanced coffee makers, automated pet feeders, and smart lighting systems.
There are robot vacuum cleaners, automated window blinds, automated dishwashers, smart ovens, self-service airport check-in kiosks, ATMs, advanced vending machines, adaptive traffic light systems, automated parking garage systems, modern HVAC systems, self-service library checkout systems, smart doorbells, advanced prosthetic limbs, automated subway or train systems, self-service bike rental kiosks, automated car washes, smart toilets, automated pill dispensers, modern pacemakers and implantable cardioverter-defibrillators, automated milking machines, self-balancing electric scooters, automated baggage handling systems, smart power grids, automated trading algorithms, and even modern hearing aids.
Our food is produced with the help of autonomous robot tractors, planting and seeding robots, weeding and pest control robots, automated greenhouse systems, harvesting robots, agricultural drones, soil analysis robots, cow-milking robots, palletizing robots, autonomous tractors, robotic sprayers, and inspection robots.
When you order a pack of batteries from Amazon.com, that order is facilitated by inventory management robots, mobile drive units for transporting goods, sorting robots, AI-driven robotic arms for item manipulation and package transfer, among others.
We are already totally reliant on robots for automotive manufacturing, electronics production, metal and machinery fabrication, plastics and chemical processing, construction, pharmaceutical manufacturing, healthcare, retail and hospitality, smart cities development, transportation, last-mile delivery, e-commerce, inspection and surveillance, and semiconductor production.
The belief that robots are always, usually, or even often humanoid is false. And the belief robots are coming in the future — well, as I just illustrated — is false, too.
We already live in an economy and society that is totally dependent upon thousands of different kinds of robots. They’re all around us, even if we don’t recognize them.
The future of robots
Meanwhile, two things are simultaneously happening. Existing jobs done by robots are constantly becoming ever more sophisticated, precise, efficient, and inexpensive.
Second, millions of tasks not previously performed by robots are being turned over to them and to robotic machinery. A diverse array of robotic devices will surely become commonplace over the next two decades.
The service sector will see a surge in delivery robots, streamlining last-mile package and food delivery logistics. Advanced cleaning robots will maintain both homes and commercial spaces.
Surgical robots performing minimally invasive procedures with high precision will benefit healthcare. Rehabilitation robots and exoskeletons will transform physical therapy and mobility, while robotic prosthetics will offer enhanced functionality to those who need them. At the microscopic level, nanorobots will revolutionize drug delivery and medical procedures.
Agriculture will increasingly embrace harvesting and planting robots to automate crop management, with specialized versions for tasks like weeding and dairy farming. Autonomous vehicles and drone delivery systems will transform the transportation sector, while robotic parking solutions will optimize urban spaces.
Military and defense applications will include reconnaissance drones, bomb disposal robots, and autonomous combat vehicles. Space exploration will continue to rely on advanced rovers, satellite-servicing robots, and assistants for astronauts on space stations.
Underwater exploration robots and devices monitoring air and water quality will benefit environmental and oceanic research. Specialized robots designed for search and rescue operations will aid disaster response efforts.
Emerging categories like soft robots, made from flexible materials, and swarm robots working collaboratively, will open new possibilities. AI-enhanced robots with advanced decision-making capabilities will become increasingly autonomous and adaptable.
In short, according to Nvidia CEO Jensen Huang (during a March 17 speech), “Everything is going to be robotic.”
When everything is robotic, what would humanoid robots do?
Musk’s prediction that humanoid robots will be teachers, babysitters, dog walkers, lawn mowers, grocery shoppers, bartenders, and friends showcases a toxic and misguided vision.
It’s misguided because nearly everything will already be robotic by the time Musk’s robots are capable of autonomous action. (To be clear, Tesla Optimus robots are nowhere near as advanced as Musk tried to make everyone believe at Tesla’s recent “We, Robot” event.
Consider, as one isolated example, the “robots will mow your lawn” claim. Musk imagines a bipedal humanoid robot in the front yard pushing a mower, an obviously ridiculous scenario.
Robotic mowers are already a banality. These appliances — sort of like Roomba vacuum cleaners, but for cutting grass instead of vacuuming floors — last for years and quietly and autonomously barber turf all day, avoiding flower beds and sidewalks without fail.
Current global sales exceed 1.5 million robotic mowers per year, at an average price of $1,200 — the same price as the cheapest Apple iPhone 16 Pro Max. Within five years, annual robotic lawn mower unit shipments are expected to reach seven million per year. Anyone with a little money and a distaste for mowing a lawn can buy one. There is no unmet need here for $50,000 humanoid robots.
This is also true for all the other jobs specified and unspecified by the Musks of the world. Robotic devices are already doing that work. So what’s left for humanoid robots to do?
The vision is toxic because, once you eliminate the tasks and jobs specialist robots will already be doing, what remains is the replacement of human connection. This very social and familial interaction is what makes us human.
Let’s start with dog walking. We get a dog so that we can enjoy the incredible connection that’s possible between a person and a dog. We feed a dog, pet a dog, hang out with a dog, and, yes, walk a dog because the experience of caring for a dog is clearly gratifying and meaningful for both species. Outsourcing our relationship with our dogs to an AI machine is (to be blunt) cold and horrible. We care for a dog for the bond and the relationship. We walk a dog in the morning for the ritual, for the exercise, and, above all, for the happiness that fulfilling the dog’s need to explore (and defecate) brings us.
Elderly companionship? Personal friendship? Education? Childcare? Why does anyone assume that a walking toaster fulfills the essential human need for empathy, connection, and care? People need to look into the sentient and conscious eyes of another human being. It’s literally the first and most basic human need.
The founder of the AI lab Midjourney, David Holz, posted on X in January that “we should be expecting a billion humanoid robots on earth in the 2040s and a hundred billion (mostly alien) robots throughout the solar system in the 2060s,” and Musk agreed with this post.
Such visionaries have what, to me, is clearly the wrong vision, and a disturbing one. We shouldn’t want AI robots to replace people in our lives and trick us into believing they’re human.
We should want non-humanoid robotic devices to do repetitive, dangerous, boring, and menial work so we can spend more time doing creative, fulfilling work and also spend more time with our parents, kids, and, yes, our pets.
Let the machines do the work. Let’s reserve our humanity for people and pets.
From its inaugural release to today, Android has transformed visually, conceptually and functionally — time and time again. Google’s mobile operating system may have started out scrappy, but holy moly, has it ever evolved.
Here’s a fast-paced tour of Android version highlights from the platform’s birth to present. (Feel free to skip ahead if you just want to see what’s new in the most recent Android 15 update.)
Android made its official public debut in 2008 with Android 1.0 — a release so ancient it didn’t even have a cute codename.
Things were pretty basic back then, but the software did include a suite of early Google apps like Gmail, Maps, Calendar, and YouTube, all of which were integrated into the operating system — a stark contrast to the more easily updatable standalone-app model employed today.
loading="lazy" width="400px">
The Android 1.0 home screen and its rudimentary web browser (not yet called Chrome).
T-Mobile
Android version 1.5: Cupcake
With early 2009’s Android 1.5 Cupcake release, the tradition of Android version names was born. Cupcake introduced numerous refinements to the Android interface, including the first on-screen keyboard — something that’d be necessary as phones moved away from the once-ubiquitous physical keyboard model.
Cupcake also brought about the framework for third-party app widgets, which would quickly turn into one of Android’s most distinguishing elements, and it provided the platform’s first-ever option for video recording.
Android 1.6, Donut, rolled into the world in the fall of 2009. Donut filled in some important holes in Android’s center, including the ability for the OS to operate on a variety of different screen sizes and resolutions — a factor that’d be critical in the years to come. It also added support for CDMA networks like Verizon, which would play a key role in Android’s imminent explosion.
loading="lazy" width="400px">
Android’s universal search box made its first appearance in Android 1.6.
Google
Android versions 2.0 to 2.1: Eclair
Keeping up the breakneck release pace of Android’s early years, Android 2.0, Eclair, emerged just six weeks after Donut; its “point-one” update, also called Eclair, came out a couple months later. Eclair was the first Android release to enter mainstream consciousness thanks to the original Motorola Droid phone and the massive Verizon-led marketing campaign surrounding it.
Verizon’s “iDon’t” ad for the Droid.
The release’s most transformative element was the addition of voice-guided turn-by-turn navigation and real-time traffic info — something previously unheard of (and still essentially unmatched) in the smartphone world. Navigation aside, Eclair brought live wallpapers to Android as well as the platform’s first speech-to-text function. And it made waves for injecting the once-iOS-exclusive pinch-to-zoom capability into Android — a move often seen as the spark that ignited Apple’s long-lasting “thermonuclear war” against Google.
loading="lazy" width="400px">
The first versions of turn-by-turn navigation and speech-to-text, in Eclair.
Google
Android version 2.2: Froyo
Just four months after Android 2.1 arrived, Google served up Android 2.2, Froyo, which revolved largely around under-the-hood performance improvements.
Froyo did deliver some important front-facing features, though, including the addition of the now-standard dock at the bottom of the home screen as well as the first incarnation of Voice Actions, which allowed you to perform basic functions like getting directions and making notes by tapping an icon and then speaking a command.
loading="lazy" width="400px">
Google’s first real attempt at voice control, in Froyo.
Google
Notably, Froyo also brought support for Flash to Android’s web browser — an option that was significant both because of the widespread use of Flash at the time and because of Apple’s adamant stance against supporting it on its own mobile devices. Apple would eventually win, of course, and Flash would become far less common. But back when it was still everywhere, being able to access the full web without any black holes was a genuine advantage only Android could offer.
Android version 2.3: Gingerbread
Android’s first true visual identity started coming into focus with 2010’s Gingerbread release. Bright green had long been the color of Android’s robot mascot, and with Gingerbread, it became an integral part of the operating system’s appearance. Black and green seeped all over the UI as Android started its slow march toward distinctive design.
loading="lazy" width="400px">
It was easy being green back in the Gingerbread days.
JR Raphael / IDG
Android 3.0 to 3.2: Honeycomb
2011’s Honeycomb period was a weird time for Android. Android 3.0 came into the world as a tablet-only release to accompany the launch of the Motorola Xoom, and through the subsequent 3.1 and 3.2 updates, it remained a tablet-exclusive (and closed-source) entity.
Under the guidance of newly arrived design chief Matias Duarte, Honeycomb introduced a dramatically reimagined UI for Android. It had a space-like “holographic” design that traded the platform’s trademark green for blue and placed an emphasis on making the most of a tablet’s screen space.
loading="lazy" width="400px">
Honeycomb: When Android got a case of the holographic blues.
JR Raphael / IDG
While the concept of a tablet-specific interface didn’t last long, many of Honeycomb’s ideas laid the groundwork for the Android we know today. The software was the first to use on-screen buttons for Android’s main navigational commands; it marked the beginning of the end for the permanent overflow-menu button; and it introduced the concept of a card-like UI with its take on the Recent Apps list.
Android version 4.0: Ice Cream Sandwich
With Honeycomb acting as the bridge from old to new, Ice Cream Sandwich — also released in 2011 — served as the platform’s official entry into the era of modern design. The release refined the visual concepts introduced with Honeycomb and reunited tablets and phones with a single, unified UI vision.
ICS dropped much of Honeycomb’s “holographic” appearance but kept its use of blue as a system-wide highlight. And it carried over core system elements like on-screen buttons and a card-like appearance for app-switching.
loading="lazy" width="400px">
The ICS home screen and app-switching interface.
JR Raphael / IDG
Android 4.0 also made swiping a more integral method of getting around the operating system, with the then-revolutionary-feeling ability to swipe away things like notifications and recent apps. And it started the slow process of bringing a standardized design framework — known as “Holo” — all throughout the OS and into Android’s app ecosystem.
Android versions 4.1 to 4.3: Jelly Bean
Spread across three impactful Android versions, 2012 and 2013’s Jelly Bean releases took ICS’s fresh foundation and made meaningful strides in fine-tuning and building upon it. The releases added plenty of poise and polish into the operating system and went a long way in making Android more inviting for the average user.
Visuals aside, Jelly Bean brought about our first taste of Google Now — the spectacular predictive-intelligence utility that’s sadly since devolved into a glorified news feed. It gave us expandable and interactive notifications, an expanded voice search system, and a more advanced system for displaying search results in general, with a focus on card-based results that attempted to answer questions directly.
Multiuser support also came into play, albeit on tablets only at this point, and an early version of Android’s Quick Settings panel made its first appearance. Jelly Bean ushered in a heavily hyped system for placing widgets on your lock screen, too — one that, like so many Android features over the years, quietly disappeared a couple years later.
Late-2013’s KitKat release marked the end of Android’s dark era, as the blacks of Gingerbread and the blues of Honeycomb finally made their way out of the operating system. Lighter backgrounds and more neutral highlights took their places, with a transparent status bar and white icons giving the OS a more contemporary appearance.
Android 4.4 also saw the first version of “OK, Google” support — but in KitKat, the hands-free activation prompt worked only when your screen was already on and you were either at your home screen or inside the Google app.
The release was Google’s first foray into claiming a full panel of the home screen for its services, too — at least, for users of its own Nexus phones and those who chose to download its first-ever standalone launcher.
loading="lazy" width="400px">
The lightened KitKat home screen and its dedicated Google Now panel.
JR Raphael / IDG
Android versions 5.0 and 5.1: Lollipop
Google essentially reinvented Android — again — with its Android 5.0 Lollipop release in the fall of 2014. Lollipop launched the still-present-today Material Design standard, which brought a whole new look that extended across all of Android, its apps and even other Google products.
The card-based concept that had been scattered throughout Android became a core UI pattern — one that would guide the appearance of everything from notifications, which now showed up on the lock screen for at-a-glance access, to the Recent Apps list, which took on an unabashedly card-based appearance.
loading="lazy" width="400px">
Lollipop and the onset of Material Design.
JR Raphael / IDG
Lollipop introduced a slew of new features into Android, including truly hands-free voice control via the “OK, Google” command, support for multiple users on phones and a priority mode for better notification management. It changed so much, unfortunately, that it also introduced a bunch of troubling bugs, many of which wouldn’t be fully ironed out until the following year’s 5.1 release.
Android version 6.0: Marshmallow
In the grand scheme of things, 2015’s Marshmallow was a fairly minor Android release — one that seemed more like a 0.1-level update than anything deserving of a full number bump. But it started the trend of Google releasing one major Android version per year and that version always receiving its own whole number.
Marshmallow’s most attention-grabbing element was a screen-search feature called Now On Tap — something that, as I said at the time, had tons of potential that wasn’t fully tapped. Google never quite perfected the system and ended up quietly retiring its brand and moving it out of the forefront the following year.
loading="lazy" width="400px">
Marshmallow and the almost-brilliance of Google Now on Tap.
JR Raphael / IDG
Android 6.0 did introduce some stuff with lasting impact, though, including more granular app permissions, support for fingerprint readers, and support for USB-C.
Android 7.0 Nougat and its new native split-screen mode.
JR Raphael / IDG
Perhaps most pivotal among Nougat’s enhancements, however, was the launch of the Google Assistant — which came alongside the announcement of Google’s first fully self-made phone, the Pixel, about two months after Nougat’s debut. The Assistant would go on to become a critical component of Android and most other Google products and is arguably the company’s foremost effort today.
Android version 8.0 and 8.1: Oreo
Android Oreo added a variety of niceties to the platform, including a native picture-in-picture mode, a notification snoozing option, and notification channels that offer fine control over how apps can alert you.
loading="lazy" width="400px">
Oreo adds several significant features to the operating system, including a new picture-in-picture mode.
JR Raphael / IDG
The 2017 release also included some noteworthy elements that furthered Google’s goal of aligning Android and Chrome OS and improving the experience of using Android apps on Chromebooks, and it was the first Android version to feature Project Treble — an ambitious effort to create a modular base for Android’s code with the hope of making it easier for device-makers to provide timely software updates.
Android version 9: Pie
The freshly baked scent of Android Pie, a.k.a. Android 9, wafted into the Android ecosystem in August of 2018. Pie’s most transformative change was its hybrid gesture/button navigation system, which traded Android’s traditional Back, Home, and Overview keys for a large, multifunctional Home button and a small Back button that appeared alongside it as needed.
Android 9 introduced a new gesture-driven system for getting around phones, with an elongated Home button and a small Back button that appears as needed.
Google released Android 10 — the first Android version to shed its letter and be known simply by a number, with no dessert-themed moniker attached — in September of 2019. Most noticeably, the software brought about a totally reimagined interface for Android gestures, this time doing away with the tappable Back button altogether and relying on a completely swipe-driven approach to system navigation.
Android 10 packed plenty of other quietly important improvements, including an updated permissions system with more granular control over location data along with a new system-wide dark theme, a new distraction-limiting Focus Mode, and a new on-demand live captioning system for any actively playing media.
Android 10’s new privacy permissions model adds some much-needed nuance into the realm of location data.
JR Raphael / IDG
Android version 11
Android 11, launched at the start of September 2020, was a pretty substantial Android update both under the hood and on the surface. The version’s most significant changes revolve around privacy: The update built upon the expanded permissions system introduced in Android 10 and added in the option to grant apps location, camera, and microphone permissions only on a limited, single-use basis.
Android 11 also made it more difficult for apps to request the ability to detect your location in the background, and it introduced a feature that automatically revokes permissions from any apps you haven’t opened lately. On the interface level, Android 11 included a refined approach to conversation-related notifications along with a new streamlined media player, a new Notification History section, a native screen-recording feature, and a system-level menu of connected-device controls.
Android 11’s new media player appears as part of the system Quick Settings panel, while the new connected-device control screen comes up whenever you press and hold your phone’s physical power button.
JR Raphael / IDG
Android version 12
Google officially launched the final version of Android 12 in October 2021, alongside the launch of its Pixel 6 and Pixel 6 Pro phones.
In a twist from the previous several Android versions, the most significant progressions with Android 12 were mostly on the surface. Android 12 featured the biggest reimagining of Android’s interface since 2014’s Android 5.0 (Lollipop) version, with an updated design standard known as Material You — which revolves around the idea of you customizing the appearance of your device with dynamically generated themes based on your current wallpaper colors. Those themes automatically change anytime your wallpaper changes, and they extend throughout the entire operating system interface and even into the interfaces of apps that support the standard.
Android 12 ushered in a whole new look and feel for the operating system, with an emphasis on simple color customization.
Google
Surface-level elements aside, Android 12 brought a (long overdue) renewed focus to Android’s widget system along with a host of important foundational enhancements in the areas of performance, security, and privacy. The update provided more powerful and accessible controls over how different apps are using your data and how much information you allow apps to access, for instance, and it included a new isolated section of the operating system that allows AI features to operate entirely on a device, without any potential for network access or data exposure.
Android version 13
Android 13, launched in August 2022, was simultaneously one of the most ambitious updates in Android history and one of the most subtle version changes to date.
On tablets and foldable phones, Android 13 introduced a slew of significant interface updates and additions aimed at improving the large-screen Android experience — including an enhanced split-screen mode for multitasking and a ChromeOS-like taskbar for easy app access from anywhere.
The new Android-13-introduced taskbar, as seen on a Google Pixel Fold phone.
Google
On regular phones, Android 13 brought about far less noticeable changes — mostly just some enhancements to the system clipboard interface, a new native QR code scanning function within the Android Quick Settings area, and a smattering of under-the-hood improvements.
Android version 14
Following a full eight months of out-in-the-open refinement, Google’s 14th Android version landed at the start of October 2023 in the midst of the company’s Pixel 8 and Pixel 8 Pro launch event.
But despite the subtle nature of its first impression, Android 14 includes a fair amount of noteworthy new stuff. The software introduces a new system for dragging and dropping text between apps, for instance, as well as a new series of native customization options for the Android lock screen.
Android 14 includes options for completely changing the appearance of the lock screen as well as for customizing which shortcuts show up on it.
JR Raphael / IDG
Android 14 provides a number of new improvements to privacy and security, too, including a new settings-integrated dashboard for managing all your health and fitness data and controlling which apps and devices can access it. And it adds in a more info-rich and context-requiring system for seeing exactly why apps want access to your location when they request such a permission.
Beyond that, Android 14 features a first taste of Google’s AI-generated custom wallpaper creator, though that’s available only on the Pixel 8 and Pixel 8 Pro to start.
You can generate all sorts of interesting wallpapers in seconds via Android 14’s AI generator feature — but only on the Pixel 8 or Pixel 8 Pro for now.
JR Raphael / IDG
The software also sports a series of significant accessibility additions, such as an enhanced on-demand magnifier, an easier way to increase font size in any app, improved support for hearing aid connections, and a built-in option to have your phone flash its camera light anytime a new notification arrives.
Android version 15
Google technically released Android 15 in September 2024, but in an unusual twist, that was only the launch of the software’s raw source code. The new Android version didn’t show up even on the company’s own Pixel devices until just over a month later, in mid-October.
As for Android 15 itself, though, the update introduces a number of noteworthy new features — including a system-level Private Space option that lets you keep sensitive apps out of sight and accessible only with authentication. The software also further enhances the multitasking systems introduced in Android 13 with the new option to keep the large-screen-exclusive Android taskbar present at all times and the new ability to launch specific pairs of apps together into a side-by-side split-screen with a single tap.
Once you set up Android 15’s new Private Space feature, certain apps appear in a special protected — and optionally hidden — area of your app drawer.
JR Raphael / IDG
Beyond that, Android 15 includes a redesigned system volume panel, a new option to automatically reenable a device’s Bluetooth radio a day after it’s been disabled, and a Pixel-specific Adaptive Vibration feature that intelligently adjusts a phone’s vibration intensity based on the environment.
Adaptive Vibration and a redesigned volume panel provide welcome upgrades to the Android audio experience.
JR Raphael / IDG
Add in a new charging-time connected-device-control screen saver, a space-saving app archiving option for infrequently used apps, and a long-under-development predictive back visual that lets you see a peek at where you’re headed before you get there — and this small-seeming update is actually shaping up to be a pretty hefty update progression.
This article was originally published in November 2017 and most recently updated in October 2024.
