Category Archives: Google

How many people are going to get the axe in favor of AI? Be Evil

Google has laid off more than a thousand people since January 10 and according to CEO Sundar Pichai, there will be more firings in the future as the company pushes forward with its AI pivot.

"We have ambitious goals and will be investing in our big priorities this year," Pichai told Google's staff in a company-wide email reviewed byThe Verge. "The reality is that to create the capacity for this investment, we have to make tough choices."

Those layoffs, as prior reporting from Business Insider andAxios indicates, took place across several departments. Hundreds were in the company's advertising and customer sales teams, while others impacted the Google Assistant, Fitbit, and hardware verticals.

In the more recent memo, the CEO said that although the more recent and forthcoming "role eliminations" will not be "at the scale of last years reductions" a reference to the 12,000 jobs Google cut around this time last year the company will continue down the path of "removing layers to simplify execution and drive velocity in some areas."

Google's forthcoming "layer-removal" comes on the heels of the company's tumultuous 2023, which along with the 12,000 cuts at the beginning of the year saw it pivot to AI much like its peers and competitors.

Indeed, as the New York Times reported based on insider interviews late last year, the company's pivot to AI was made hastily as OpenAI's ChatGPT and its alliance with Microsoft threatened to leave more established tech players else in its dust.

Just before Christmas 2022, theNYT's reporting explains, the company's top lawyer summoned executives to deliver a directive: that their teams were to drop whatever they had previously been working on and begin developing a slate of AI products immediately, per Pichai's orders. Just over a month later, those 12,000 employees were axed, and now we appear to be seeing the continued toll of that rapid decision-making process.

AsAxios aptly pointed out in its own analysis of the 2024 Google layoffs, it seems less that people are being replaced by AI itself and more that they are being replaced with smaller teams of people who are good with the technology as the company adjusts course.

It's hard to say how those AI-adept new workers feel about their roles, but it seems likely they're essentially building the tech that will eventually replace them and that can't feel good.

More on Google: Image Database Powering Google's AI Contains Explicit Images of Children

See the original post:

As Google Pivots to AI, CEO Warns He Will Fire Even More Staff - Futurism

Posted by Jiefeng Chen, Student Researcher, and Jinsung Yoon, Research Scientist, Cloud AI Team

In the fast-evolving landscape of artificial intelligence, large language models (LLMs) have revolutionized the way we interact with machines, pushing the boundaries of natural language understanding and generation to unprecedented heights. Yet, the leap into high-stakes decision-making applications remains a chasm too wide, primarily due to the inherent uncertainty of model predictions. Traditional LLMs generate responses recursively, yet they lack an intrinsic mechanism to assign a confidence score to these responses. Although one can derive a confidence score by summing up the probabilities of individual tokens in the sequence, traditional approaches typically fall short in reliably distinguishing between correct and incorrect answers. But what if LLMs could gauge their own confidence and only make predictions when they're sure?

Selective prediction aims to do this by enabling LLMs to output an answer along with a selection score, which indicates the probability that the answer is correct. With selective prediction, one can better understand the reliability of LLMs deployed in a variety of applications. Prior research, such as semantic uncertainty and self-evaluation, has attempted to enable selective prediction in LLMs. A typical approach is to use heuristic prompts like Is the proposed answer True or False? to trigger self-evaluation in LLMs. However, this approach may not work well on challenging question answering (QA) tasks.

In "Adaptation with Self-Evaluation to Improve Selective Prediction in LLMs", presented at Findings of EMNLP 2023, we introduce ASPIRE a novel framework meticulously designed to enhance the selective prediction capabilities of LLMs. ASPIRE fine-tunes LLMs on QA tasks via parameter-efficient fine-tuning, and trains them to evaluate whether their generated answers are correct. ASPIRE allows LLMs to output an answer along with a confidence score for that answer. Our experimental results demonstrate that ASPIRE significantly outperforms state-of-the-art selective prediction methods on a variety of QA datasets, such as the CoQA benchmark.

Imagine teaching an LLM to not only answer questions but also evaluate those answers akin to a student verifying their answers in the back of the textbook. That's the essence of ASPIRE, which involves three stages: (1) task-specific tuning, (2) answer sampling, and (3) self-evaluation learning.

Task-specific tuning: ASPIRE performs task-specific tuning to train adaptable parameters (p) while freezing the LLM. Given a training dataset for a generative task, it fine-tunes the pre-trained LLM to improve its prediction performance. Towards this end, parameter-efficient tuning techniques (e.g., soft prompt tuning and LoRA) might be employed to adapt the pre-trained LLM on the task, given their effectiveness in obtaining strong generalization with small amounts of target task data. Specifically, the LLM parameters () are frozen and adaptable parameters (p) are added for fine-tuning. Only p are updated to minimize the standard LLM training loss (e.g., cross-entropy). Such fine-tuning can improve selective prediction performance because it not only improves the prediction accuracy, but also enhances the likelihood of correct output sequences.

Answer sampling: After task-specific tuning, ASPIRE uses the LLM with the learned p to generate different answers for each training question and create a dataset for self-evaluation learning. We aim to generate output sequences that have a high likelihood. We use beam search as the decoding algorithm to generate high-likelihood output sequences and the Rouge-L metric to determine if the generated output sequence is correct.

Self-evaluation learning: After sampling high-likelihood outputs for each query, ASPIRE adds adaptable parameters (s) and only fine-tunes s for learning self-evaluation. Since the output sequence generation only depends on and p, freezing and the learned p can avoid changing the prediction behaviors of the LLM when learning self-evaluation. We optimize s such that the adapted LLM can distinguish between correct and incorrect answers on their own.

In the proposed framework, p and s can be trained using any parameter-efficient tuning approach. In this work, we use soft prompt tuning, a simple yet effective mechanism for learning soft prompts to condition frozen language models to perform specific downstream tasks more effectively than traditional discrete text prompts. The driving force behind this approach lies in the recognition that if we can develop prompts that effectively stimulate self-evaluation, it should be possible to discover these prompts through soft prompt tuning in conjunction with targeted training objectives.

After training p and s, we obtain the prediction for the query via beam search decoding. We then define a selection score that combines the likelihood of the generated answer with the learned self-evaluation score (i.e., the likelihood of the prediction being correct for the query) to make selective predictions.

To demonstrate ASPIREs efficacy, we evaluate it across three question-answering datasets CoQA, TriviaQA, and SQuAD using various open pre-trained transformer (OPT) models. By training p with soft prompt tuning, we observed a substantial hike in the LLMs' accuracy. For example, the OPT-2.7B model adapted with ASPIRE demonstrated improved performance over the larger, pre-trained OPT-30B model using the CoQA and SQuAD datasets. These results suggest that with suitable adaptations, smaller LLMs might have the capability to match or potentially surpass the accuracy of larger models in some scenarios.

When delving into the computation of selection scores with fixed model predictions, ASPIRE received a higher AUROC score (the probability that a randomly chosen correct output sequence has a higher selection score than a randomly chosen incorrect output sequence) than baseline methods across all datasets. For example, on the CoQA benchmark, ASPIRE improves the AUROC from 51.3% to 80.3% compared to the baselines.

An intriguing pattern emerged from the TriviaQA dataset evaluations. While the pre-trained OPT-30B model demonstrated higher baseline accuracy, its performance in selective prediction did not improve significantly when traditional self-evaluation methods Self-eval and P(True) were applied. In contrast, the smaller OPT-2.7B model, when enhanced with ASPIRE, outperformed in this aspect. This discrepancy underscores a vital insight: larger LLMs utilizing conventional self-evaluation techniques may not be as effective in selective prediction as smaller, ASPIRE-enhanced models.

Our experimental journey with ASPIRE underscores a pivotal shift in the landscape of LLMs: The capacity of a language model is not the be-all and end-all of its performance. Instead, the effectiveness of models can be drastically improved through strategic adaptations, allowing for more precise, confident predictions even in smaller models. As a result, ASPIRE stands as a testament to the potential of LLMs that can judiciously ascertain their own certainty and decisively outperform larger counterparts in selective prediction tasks.

In conclusion, ASPIRE is not just another framework; it's a vision of a future where LLMs can be trusted partners in decision-making. By honing the selective prediction performance, we're inching closer to realizing the full potential of AI in critical applications.

Our research has opened new doors, and we invite the community to build upon this foundation. We're excited to see how ASPIRE will inspire the next generation of LLMs and beyond. To learn more about our findings, we encourage you to read our paper and join us in this thrilling journey towards creating a more reliable and self-aware AI.

We gratefully acknowledge the contributions of Sayna Ebrahimi, Sercan O Arik, Tomas Pfister, and Somesh Jha.

View original post here:

Introducing ASPIRE for selective prediction in LLMs Google Research Blog - Google Research

With the new icon widely rolled out, Google Chats homescreen redesign is now making its way to Android.

Update 1/17/24: Were now seeing much wider availability of the redesigned Google Chat. The rollout has ramped up in recent days and we now have it on all our devices after force closing the standalone Chat app and Gmail. (In the latter app, the homescreen widget has also been updated to just feature two tabs in the pseudo-bottom bar.)

You still have to manually enable the Chat tab from Gmail Settings > your email address > General. Unlike Meet, Google never rolled out the integrated Google Chat experience to all users.

Original 12/19/23: In the standalone Google Chat app, the two-tab bottom bar and New FAB as seen below have been replaced.

Instead of tabs for Chat and Spaces, you now have Home, Direct messages, Spaces, and Mentions in a pill-shaped container.

A circular indicator notes what tab youre currently viewing. (A nice touch would have been having the circle glide/animate through as you switch between feeds.)There are no labels underneath each icon, but you can find that information at the top of your screen below the search bar. Thankfully, the Google Chat icons are pretty straightforward.

Meanwhile, theres a rounded square FAB to the right of the pill that is ever so slightly smaller. These two elements are centered.

In Gmail, this floating pill appears above the existing bottom bar, which has been consolidated to three tabs. While this navigation approach is fine in the dedicated app, the integrated Gmail experience looks rather busy.

This redesign first rolled out to Google Chat and Gmail for iOS, as well as the web. On Android, its slowly appearing on an account-by-account basis with a server-side update. Were seeing it on a personal Google Account today.

FTC: We use income earning auto affiliate links. More.

Original post:

New Google Chat redesign widely rolling out on Android [U] - 9to5Google

Google pledged to stop tracking user visits to abortion clinics shortly after Roe v. Wade was overturned in June 2022, which killed off the United States largest federal abortion protection. The measure aimed to protect abortion seekers from prosecutors, especially in the 14 states where abortion is now banned. Roughly 18 months later, Google has not followed through on its promise, according to a new study from Accountable Tech, and the company still tracks visits to abortion clinics.

No Google AI Search, I Dont Need to Learn About the Benefits of Slavery

Googles location data is one of law enforcements favorite tools in the United States. The U.S. government requested the companys data over 63,000 times in the first half of 2023, reaching an all-time high, and Google handed over data 85% of the time. Thats why the companys tracking of visits to abortion clinics is a big deal, and why its promise to stop was so celebrated. However, a new study published on Thursday ran tests in seven states across the country and found that Google is still collecting Location History data for visits to abortion clinics, despite their promise. Accountable Tech says this report makes it clear that Google cannot be trusted to follow-through on its privacy promises.

The study ran experiments in Pennsylvania, Texas, Nevada, Florida, New York, Georgia, and North Carolina. In 50% of its tests, Accountable Tech found that routes to and from Planned Parenthood locations were saved in Googles Location History. The actual name, Planned Parenthood, was scrubbed from Googles data, which mirrors Accountable Techs study from a year ago, but the data still clearly points to abortion clinic visits.

Google announced last December it would no longer hold onto peoples location data which would mean it cant turn that information over to the police. Accountable Tech says this was a positive development, but that Google cant be trusted at face value on promises like this. The company received 5,764 geofence warrants in states where abortion is banned between 2018 and 2020, according to Politico.

Google did not immediately respond to Gizmodos request for comment.

Google notes that its Location History feature is default turned off, and you can turn it on if you choose. Its unclear how many users Google tracks location data on, but its estimated the company is one of the worlds largest data trackers, with over a billion users of Google Maps globally.

Follow this link:

Google Breaks Another Promise About Tracking Your Location History - Gizmodo

Google searches on high-end Android phones are going to get substantially more powerful, thanks to the new Circle to Search feature, the company said in a press releaseWednesday. The feature lets you pull up Google while using any app; circle or highlight any text or object; and, with the power of AI, instantly get search results about that text or object. This enhanced version of visual search will launch on the Pixel 8, Pixel 8 Pro, and theSamsung Galaxy S24 line of phones on Jan. 31.

Circle to Search, as the name suggests, lets you circle any object to search for it. For example, if you see a TikTok featuring funky-looking corn dogs, you can swipe up to bring up the Google app and circle the food item with your finger. Using AI, Google will give you the brief on Korean corn dogs. The AI can also explain the cultural relevance of Korean food in the current gastronomical zeitgeist.

The power here isn't only being able to search any object or text you see on your screen but doing so without having to fumble between multiple apps.

This announcement coincides with Samsung's announcement today of the same feature coming to the new Galaxy S24 line of phones. At the moment, other phones with this more powerful in-app version of Google Search haven't been announced.

Watch this: 'Circle to Search' Lets Users Google From Any Screen

Google's continued love affair with AI comes after ChatGPT upended its Search mojo in late 2022. When ChatGPT launched, the AI chatbot could seemingly answer any question with unique answers. Compared with a standard Google search, which brings up lists of websites requiring you to filter and find relevant information, ChatGPT did that for you. Both Google and Microsoft were quick to introduce AI chatbots of their own to stay competitive.

Since then, Microsoft's AI investments has catapulted the company into becoming the most valuable in the world, overtaking Apple earlier this month. Google is also integrating AI into more of its products, from Search itself withSearch Generative Experienceto the camera app on its Pixel devices. Considering that generative AI is estimated to bring in up to $4.4 trillion annually, according to McKinsey, companies are rushing to be at the cutting edge to maintain market dominance.

Along with Circle to Search, Google said that multisearch in Lens is also getting an AI-powered upgrade. In the past, if you saw a dress you liked, Google Lens would be able to identify that dress and you could then ask Google to search for variations in green. Now those visual matches will give AI-powered insights, allowing you to ask more nuanced questions. The example Google posits is seeing a board game at a yard sale without the box. Google Lens can soon identify the game and you can ask what that game is called and how it can be played.

Editors' note: CNET is using an AI engine to help create some stories. For more, seethis post.

Read the original here:

Samsung Galaxy S24 Uses Google's Circle to Search to Look Up Anything on Your Phone's Screen - CNET

After laying off about 1,000 employees last week, Google is now cutting jobs on its advertising sales team. In a statement to The Verge, Google spokesperson Chris Pappas confirmed that a few hundred roles globally are being eliminated as part of the change.

The news of Googles latest layoffs was first reported by Business Insider, which obtained a memo from Google senior vice president Philipp Schindler. Schindler reportedly hinted the layoffs will primarily affect its Large Customer Sales (LCS) unit, a team that sells ads to large businesses, while the Google Customer Solutions team (GCS), which sells ads to smaller clients, will become the core ad sales team instead. Google laid off some employees on its LCS team last October, according to a separate report from Business Insider.

Every year we go through a rigorous process to structure our team to provide the best service to our Ads customers, Pappas tells The Verge. We map customers to the right specialist teams and sales channels to meet their service needs.As part of this, a few hundred roles globally are being eliminated and impacted employees will be able to apply for open roles or elsewhere at Google.

Google still hasnt put a number on how many employees it has laid off this year. But with last weeks layoffs affecting a few hundred roles in each impacted division and this most recent round eliminating another few hundred jobs, the cuts have certainly climbed above 1,000.

Follow this link:

Google layoffs continue with hundreds from sales team - The Verge

Google will let global Android smartphone users search for content in other applications without having to exit them, a move aimed at boosting its search engine business and driving engagement for app developers.

The company's new Circle to Search feature, to be rolled out gradually from January 31 on some premium Android smartphones, will allow users to select text, images or videos in any other apps, and have Google search them within the apps, the US technology giant said on Thursday.

For example, users will be able to search for details on what a content creator is wearing in a video by circling or tapping on the items, after they activate the Circle to Search function by long-pressing their Android phones home button or navigation bar, the company said.

The feature will enable quick access to information without having to leave the app, potentially driving engagement for apps and the content displayed, it said.

Apple gets deeper into Chinas smartphone price wars with broad iPhone discounts

Early testing showed that users typically return to the app theyre using after activating the feature, according to Google.

The company did not disclose when the function will be available on iPhones, with a representative telling the South China Morning Post that it looks forward to bringing this powerful new capability to more people over time.

Android handsets that will feature the new function will include Samsung Electronics new Galaxy S24 Series, it said.

Google parent Alphabets revenue in the three months ending September 30 rose 11 per cent from the same period a year before, which the company attributed to meaningful growth in search and YouTube, according to its earnings report published in October. Advertising revenue grew 9.5 per cent year on year to US$59.6 billion.

The company is still facing uncertainties in Hong Kong, where the government is seeking to have the US technology giant remove the protest song Glory to Hong Kong, which became popular during anti-government unrest in the city in 2019.

Google has expressed unwillingness to ban the song from its search engine without a court order.

Follow this link:

Google to let global Android users search for content within other apps - South China Morning Post

Mirroring Meta, Google announced a series of layoffs impacting the augmented reality (AR) hardware, Google Assistant, and core engineering divisions, last week (Jan. 10).

A Google representative explained to Android Central that these changes were geared towards enhancing efficiency and aligning resources with the company's major product priorities.

However, according to an internal memo obtained by The Verge, Google CEO Sundar Pichai told employees on Wednesday (Jan. 17) that the layoffs this year focused on removing layers to simplify execution and drive velocity in some areas.

We have ambitious goals and will be investing in our big priorities this year, Pichai wrote.

The reality is that to create the capacity for this investment, we have to make tough choices."

He acknowledged the changes faced by several teams and hinted that more layoffs were making their way.

However, Pichai assured his employees that these role eliminations are not at last year's scale of reductions and will not impact every team.

Google's parent company, Alphabet, laid off 12,000 jobs, 6% of its employees, in January 2023 because they over-hired people to match and fuel two years of dramatic growth.

"But I know it's difficult to see colleagues and teams impacted," Pichai noted in the memo.

It is speculated that these layoffs are occurring due to Google's recent shift towards AI. Google toldSemafor, "restructuring would help improve Google Assistant as it explores integrating newer artificial intelligence technology into its products."

According to the Alphabet Workers Union, Google laid off nearly 15,000 employees over the past year.

"Many of these changes are already announced, though to be upfront, some teams will continue to make specific resource allocation decisions throughout the year where needed, and some roles may be impacted," Pichai added.

Not much was said about which departments would be impacted or when the layoffs were expected to happen.

Google has assured affected employees of its support and urged them to seek new opportunities within the company. Nevertheless, given the recent layoffs and more anticipated, prospects of rehiring seem bleak.

Read more:

More job cuts on the horizon for Google says CEO Sundar Pichai - Android Central

Microsoft and Google have teamed up to make the default Edge web browser in Windows 11 feel a lot faster and smoother to use and the tech used could benefit Chrome users as well.

As the Winbuzzer website has noticed, this is all thanks to Windows 11s Application Launch Prefetch range (APLF), which has been expanded, potentially boosting its performance.

APLF comes with a parameter known as /prefetch which instructs how Windows 11 optimizes and loads applications. In short, this means Windows 11 can prioritize certain programs and dedicate more processing power or memory to the app and that could mean Edge will load faster when you start it up, display websites (especially image-heavy ones) more smoothly, and just generally feel a lot quicker.

While it wont magically make your internet connection faster, it could still make the process of browsing the web in Edge feel a lot snappier and responsive. The increased range of the /prefetch parameter, highlighted in a code submission to the Chromium project, will allow for better process separation, especially for renderers, and utility processes, so Windows 11 will be better at optimizing resources for Edge.

The fact that Google has been lending a hand to improve the performance of Microsofts Edge browser may seem a little bit odd since the search giant has its own rival browser Chrome.

However, this collaboration makes a lot more sense if you take into account that both Edge and Chrome run on the same underlying Chromium code so by helping improve Edge performance in Windows 11, Google should also see those improvements coming to Chrome as well.

This is the beauty of big companies like Google and Microsoft working on open-source projects like Chromium, as it means those projects can get the benefit of features and fixes from teams that have the resources, training, and experience that might otherwise be kept in-house.

At the moment, improved Prefetch compatibility is included in the latest Edge Canary build, which is an early version used by developers to test out new features. If it works as expected, it should then be included in an update for the Edge browser everyone uses.

APLF is also slated to be made available for any application that runs in Windows 11, so this work wont just benefit Chrome, Edge, and other Chromium-based browsers either. Exciting stuff!

Read more:

Microsoft Edge gets a speed boost in Windows 11 with a little help from Google - TechRadar

There were some huge AI announcements at CES 2024 in Las Vegas last week, but it was Google killing off over a dozen features of its Assistant that stuck with me well after the show. The company deems this action necessary to streamline its workflow and free up resources going towards maintaining "underutilized" services. I'm seeing a more worrisome picture. With more than hundreds of recent layoffs in Google's voice assistant and hardware divisions, Google Assistant appears to be in a rocky place.

I say that as someone who has been using the platform almost daily since the 2016 release of the original Google Home smart speaker. Although my hardware has evolved to include the latest Google Nest Hub (2nd Gen) and Pixel Tablet, I've noticed a gradual decline in Google Assistant's speed and voice command accuracy over time. It's not just an issue on my end (I've troubleshot my network and speak clearly) pop over to threads on social platforms like Reddit or X and you'll see others complain and even post memes about common mishaps.

Between declining performance and Google removing core features like the ability to set an alarm to music or send an email by voice, Assistant is painting itself into a corner of becoming less useful. All 17 features being removed are listed below and can also be found on Google's support site.

These changes hit especially hard in light of Amazon Alexa's growing functionality. Amazon debuted its open Matter casting protocol which isn't locked to specific hardware. Unlike Google's Chromecast you'll be able to beam content up to even more displays and devices going forward. Alexa's new generative AI-powered experiences allow for apps like Character.AIwhich lets you have full-on conversations and get advice from profiles that range from life coaches to historical figures. And if you have an Alexa-controlled smart home you can use Map View to create a 3D floorplan of it and place your devices on the map as interactive pins. This lets you instantly tap on and control a device without sorting through a list like in Google Home which lacks a competing visual interface.

All of these capabilities only make up what's brand new. The Alexa platform has been steadily improving over time. In addition to performance refinements, Amazon has added new triggers like the "listen for beep" function that made it easy to turn my old washing machine smart. I'm not repeating commands anywhere near as frequently as I am with Google. Plus my Alexa devices can still show me step-by-step recipes, play my Kindle audiobooks, and set media alarms just fine.

Theres speculation that Google Assistant's feature cuts may be a potential first step for the tech giant to implement better AI across its platform. At its Pixel 8 event late last year, the company announced plans to launchan upgraded Google Assistant with Bard. Bard is the brand's generative AI that can create personalized answers based on events, dates, and conversations stored on your devices and Google account. Beyond voice commands, Google Assistant with Bard's contextual awareness can take actions like sort through and summarize your emails or even make you dinner reservations. However, Google didn't say if that version has anything to do with the impacts on current Assistant functionality.

Google Assistant is powerful with more contextual knowledge and a great UI (both visual and voice), but worse than not making headway with the addition of major features or improving upon the experience as others have, it's actively regressing. Should Google expand its features with an upgraded generative AI assistant it can earn back a spot in my main living areas which are dominated by Alexa. For now, its smart displays sit in my bedroom and office.

Unfortunately, the functions being cut are a big part of why those devices ended up in these spots. I use the Nest in my bedroom because it doesn't have a built-in camera, can track my sleep without a wearable, and wake me up to music on its alarm. So much for the sleep summaries and media alarm. And in my office, I occasionally use the Assistant with my personal Gmail account. After next month they'll mainly just be fancy clocks.

Today's best Google Nest Hub (2nd Gen) deals

The rest is here:

Google just gave me 17 more reasons why I'm sticking with Alexa - Tom's Guide