Рубрика: AI

To deliver the high demand for HPC, Microsoft announced its intention to provide its cloud computing clients with AI chips developed by AMD.

Additionally, Microsoft will unveil a preview of its new Cobalt 100 custom processors at the upcoming conference. These chips, first announced in November, reportedly deliver a 40% performance improvement over other ARM-designed competitors. Companies such as Adobe and Snowflake are already utilising these processors.

During a keynote speech, Microsoft introduced the Copilot+ PC, their brand of AI PCs. This sets specific hardware requirements for systems to qualify as ‘AI-first’ PCs. These PCs will have the capability to manage certain AI-accelerated tasks, such as chatbots and image generation locally, rather than depending on cloud services.

However, new hardware will be necessary to execute these tasks swiftly and efficiently. For this, Microsoft has announced a partnership with Intel, AMD, and Qualcomm to build neural processing units (NPUs), allowing these AI models to run locally on PCs.

Minimum requirements built for Qualcomm

At a minimum, systems will need 16GB of RAM and 256GB of SSD to meet the memory and on-disk storage requirements for LLMs such as Microsoft’s Phi-3. Microsoft has announced that all PCs powered by the Snapdragon X Plus and X Elite chips will have Copilot+ features pre-installed. According to Microsoft, they will start shipping on June 18th.

During the keynote speech, Microsoft said that the Copilot+ PC runs 58% faster than the MacBook Air M3 in sustained multi-threaded performance on Snapdragon X Elite.

All this sounds fine for now for Intel and AMD.

However, the most significant new requirement, which will pose a challenge for almost all existing Windows PCs, is the need for an integrated NPU. Microsoft mandates an NPU with a performance rating of 40 trillion operations per second (TOPS), a benchmark used by Microsoft, Qualcomm, Apple, and others in comparing NPU capabilities.

The challenge for Intel and AMD is that none of their current-generation chips meet this requirement, even those equipped with NPUs. Intel’s Meteor Lake-based Core Ultra NPUs peak at 10 TOPS, while some of AMD’s Ryzen 7000 and Ryzen 8000 desktop and laptop processors with NPUs range between 12 and 16 TOPS.

Currently, this requirement is met by only one chip in the Windows PC ecosystem: Qualcomm’s Snapdragon X Elite and X Plus. These chips, set to launch in the new Surface PCs and offerings from Dell, Lenovo, HP, Asus, Acer, and other major OEMs in the coming months, feature NPUs capable of 45 TOPS, slightly exceeding Microsoft’s minimum requirement.

In comparison, Apple’s latest M4 processor is capable of 38 TOPS of NPU performance.

While both Intel and AMD have future products in their roadmap aimed at enhancing NPU performance to meet Microsoft’s Copilot+ PC requirements, none of the existing hundreds of millions of x86-based Windows PCs, including current models available for purchase, will qualify for the Copilot+ label.

Intel and AMD’s CPUs and GPUs can add to the total TOPS a system can achieve. However, NPUs are specialised hardware designed to execute AI workloads more efficiently and with lower power consumption.

Intel and AMD Catching up

As mentioned before, both Intel and AMD have plans in the pipeline to meet these requirements. Intel has revealed that starting from the third quarter of 2024, its highly anticipated client processors, codenamed Lunar Lake, are slated to power over 80 fresh laptop designs across more than 20 OEMs.

Lunar Lake is boasting over three times the AI performance of its predecessors. With an impressive 40+ NPU TOPS, Intel’s next-gen processors are poised to deliver the capabilities required for the upcoming Copilot+ experiences. Moreover, Lunar Lake will feature over 60 GPU TOPS, amounting to more than 100 platform TOPS in total.

Similarly, AMD is also coming up with its new APU, the Ryzen 8050, featuring a Zen 5 CPU and an XDNA 2 NPU architecture for AI PC workloads. This will boast a performance of around 50 TOPS, more than ideal for running Microsoft’s AI PC goals. These have been dubbed the Strix Point APUs. AMD is expected to launch these by the second half of 2024.

Once released, the two will try to outcompete each other as to who can offer better AI PC goals. Meanwhile, Qualcomm, with its Snapdragon X Plus and X Elite chips, will start shipping by the end of next month.

This is all while NVIDIA criticises Microsoft’s AI PC conversations. The company said that its GPUs provide better performance than any NPUs in the market, calling all other NPUs only good enough for ‘basic’ AI tasks. It also claims that its RTX GPUs can achieve between 100 to 1,300 TOPS of power, depending on the GPU.

The post Microsoft Shows Love for Intel and AMD But Chooses Qualcomm For Copilot+ PCs appeared first on AIM.

Generative artificial intelligence. GenAI for short. It’s the one topic that has companies excited more than anything else in the past few years. And for good reason; it pulls the best information possible out of your data. It has the potential to transform how a business operates, how employees work, and how we all live our daily lives. It’s a powerful tool that has placed the world on the edge of a new sort of Industrial Revolution. For many companies, it’s a brand-new territory. At Axtria, we’ve been working with data analytics across our products and solutions since day one. Uncovering hidden insights from information – what GenAI does best – is part of our DNA.

What about the companies that don’t have data so naturally embedded in their culture? How can they join the GenAI party? They can start by following four key concepts crucial to all GenAI strategies, irrespective of industry.

1. Choose GenAI use cases that provide measurable impact. Some companies take a shotgun approach: trying GenAI in everything and seeing what works. As you can imagine, that’s a massive waste of resources. Maybe you want to improve a customer’s buying journey. Perhaps you want your sales reps to get insights and tips before walking into a meeting. Remember, not all use cases will have the same level of organizational impact, so choose wisely. You need someone to help you tailor a GenAI journey for specific projects that are scalable and will ensure adoption by your team members. Consider working with an established firm that can guide you and enable your GenAI journey of exploration, experimentation, industrialization, adoption, and, sometimes, monetization.

1. Build strong Generative Relational Databases. You’ll quickly realize that every bit of data you have holds value, even the type that doesn’t fit neatly into a spreadsheet. “Traditional” data—names, numbers, sales figures, and such—needs to be clean and error-free. GenAI can parse the not-so-neatly-organized data—photos, video, audio, even handwritten notes. Combine it all by leveraging what are called Generative-Ready Datasets (GRD). These GRDs train the Large Language Models that form the basis of GenAI. This is important: Get this correct at the start, and you won’t have to worry about mistakes scaling with you. Your goal is to industrialize your GenAI usage, and this GRD strategy sets you up for it, not just with larger datasets but across all business functions. The potential is there for the taking if you seek products and platforms that leverage GenAI in production.

1. Ensure a viable, intertwined GenAI and AI strategy. You can explore and experiment with GenAI, but industrialization can only happen once you have validated the strategy – by accepting the proofs-of-concept. That means you have to sit down and have frank discussions on sensitive issues: intellectual property protection, personal data, and other critical areas across your products and solutions. Setting the proper guardrails here can avoid severe reputational damage and prevent stumbles later on. Relying on a trusted and experienced partner who can help you look around corners is essential.

1. Define the characteristics of industrialization excellence. In order to scale up, you need to have a platform that people are willing to use. The user interface must be simple to navigate, with relevant information easy to find. The user experience has to be positive, or you’ll never get full adoption. Likewise, you need to ensure your business users are on board with a new platform. Getting buy-in and positive change management ensures trust in the AI models. If your team isn’t using the model, or they don’t trust it, it’s meaningless. Once you have that in place, you’re now ready to start turning promises into wider practice. Find a partner who listens to your industrialization goals and has expertise with products and platforms that leverage GenAI in production. One who can help you define the best practices and what will be considered a success. Remember, industrialization with GenAI doesn’t have to be limited to “more customers.” GenAI can find efficiencies in your own team, so don’t forget to look inwards.

When you’ve got your GenAI strategy humming along beautifully, you must remain vigilant. Responsible GenAI usage involves security, as well as frequent monitoring of the analysis models, user controls, and usage compliance. Accountability is a crucial factor as well. You must document everyone tasked with ownership, oversight, and approval at every stage, including the usage of third-party models.

Another critical aspect of GenAI responsibility comes from the human factor. You have to commit to eliminating biases in model training. We all have opinions, and that makes us unique. Be sure those feelings don’t handcuff the model, leaving important insights uncovered.

Finally, you have to be able to explain your GenAI model. It must be transparent and traceable, which helps with governance considerations. Seeing how a model developed its answer helps with model refinement and reporting responsibilities.

Putting these all together will give you a GenAI strategy that’s effective and well-suited for industrialization. It’s a big ask of any company, but leaning on a partner with established know-how across products and solutions will get you up and running faster and safer than doing it yourself. Since our founding in 2010, Axtria has focused on helping our clients make better decisions by enabling the best use of data. Our life sciences clients, whose work helps save lives, depend on Axtria and our rich expertise in leveraging data to help them work better, smarter, and with greater impact.

Interested to read more on the impact and future of GenAI:

• Reading the Tea Leaves: New Insights from Gartner® on the Future of Gen AI and the Value it’s Already Adding in Life Sciences
• The Use of Natural Language Processing in Literature Reviews

The post Generative AI: How to Move from Promises to Production appeared first on AIM.

Image by Author | DALLE-3 & Canva

Data Engineering is rapidly growing, and companies are now hiring more data engineers than data scientists. Operational jobs like data engineering, cloud architecture, and MLOps engineering are in high demand.

As a data engineer, you need to master containerization, infrastructure as code, workflow orchestration, analytical engineering, batch processing, and streaming tools. Apart from these tools, you need to master cloud infrastructure and manage services like Databricks and Snowflakes.

In this blog, we will learn about 10 GitHub repositories that will help you master all core tools and concepts. These GitHub repositories contain courses, experiences, roadmaps, a list of essential tools, projects, and a handbook. All you need to do is bookmark them while learning to become a professional data engineer.

1. Awesome Data Engineering

The Awesome Data Engineering repository contains a list of tools, frameworks, and libraries for data engineering, making it an excellent starting point for anyone looking to dive into the field.

It covers tools on databases, data ingestion, files system, streaming, batch processing, data lake management, workflow orchestration, monitoring, testing, and charts and dashboards.

Link: igorbarinov/awesome-data-engineering

2. Data Engineering Zoomcamp

Data Engineering Zoomcamp is a complete course that provides a hands-on learning experience in data engineering. You learn new concepts and tools using video tutorials, quizzes, projects, homework, and community-driven assessments.

The Data Engineering Zoomcamp covers:

1. Containerization and Infrastructure as Code
2. Workflow Orchestration
3. Data Ingestion
4. Data Warehouse
5. Analytics Engineering
6. Batch processing
7. Streaming

Link: DataTalksClub/data-engineering-zoomcamp

3. The Data Engineering Cookbook

The Data Engineering Cookbook is a collection of articles and tutorials that cover various aspects of data engineering, including data ingestion, data processing, and data warehousing.

The Data Engineering Cookbook includes:

1. Basic Engineering Skills
2. Advanced Engineering Skills
3. Free Hands On Courses / Tutorials
4. Case Studies
5. Best Practices Cloud Platforms
6. 130+ Data Sources Data Science
7. 1001 Interview Questions
8. Recommended Books, Courses, and Podcasts

Link: andkret/Cookbook

4. Data Engineer Roadmap

The Data Engineer Roadmap repository provides a step-by-step guide to becoming a data engineer. This repository covers everything from the basics of data engineering to advanced topics like Infrastructures as a code and cloud computing.

The Data Engineer Roadmap includes:

1. CS fundamentals
2. Learning Python
3. Testing
4. Database
5. Data Warehouse
6. Cluster Computing
7. Data Processing
8. Messaging
9. Workflow Scheduling
10. Network
11. Infrastructures as a Code
12. CI/CD
13. Data Security and Privacy

Link: datastacktv/data-engineer-roadmap

5. Data Engineering HowTo

Data Engineering HowTo is a beginner-friendly resource for learning data engineering from scratch. It contains a list of tutorials, courses, books, and other resources to help you build a solid foundation in data engineering concepts and best practices. If you're new to the field, this repository will help you navigate the vast landscape of data engineering with ease.

How To Become a Data Engineer includes:

1. Useful articles and blogs
2. Talks
3. Algorithms & Data Structures
4. SQL
5. Programming
6. Databases
7. Distributed Systems
8. Books
9. Courses
10. Tools
11. Cloud Platforms
12. Communities
13. Jobs
14. Newsletters

Link: adilkhash/Data-Engineering-HowTo

6. Awesome Open Source Data Engineering

Awesome Open Source Data Engineering is a list of open-source data engineering tools that is a goldmine for anyone looking to contribute to or use them to build real-world data engineering projects. It contains a wealth of information on open-source tools and frameworks, making it an excellent resource for anyone looking to explore alternative data engineering solutions.

The repository includes open-source tools on:

1. Analytics
2. Business Intelligence
3. Data Lakehouse
4. Change Data Capture
5. Datastores
6. Data Governance and Registries
7. Data Virtualization
8. Data Orchestration
9. Formats
10. Integration
11. Messaging Infrastructure
12. Specifications and Standards
13. Stream Processing
14. Testing
15. Monitoring and Logging
16. Versioning
17. Workflow Management

Link: gunnarmorling/awesome-opensource-data-engineering

7. Pyspark Example Project

Pyspark Example Project repository provides a practical example of implementing best practices for PySpark ETL jobs and applications.

PySpark is a popular tool for data processing, and this repository will help you master it. You will learn how to structure your code, handle data transformations, and optimize your PySpark workflows efficiently.

The project covers:

1. Structure of an ETL Job
2. Passing Configuration Parameters to the ETL Job
3. Packaging ETL Job Dependencies
4. Running the ETL job
5. Debugging Spark Jobs
6. Automated Testing
7. Managing Project Dependencies

Link: AlexIoannides/pyspark-example-project

8. Data Engineer Handbook

Data Engineer Handbook is a comprehensive collection of resources covering all aspects of data engineering. It includes tutorials, articles, and books on all the topics related to data engineering. Whether you are looking for a quick reference guide or in-depth knowledge, this handbook has something for data engineers of all levels.

The Handbook includes:

1. Great Books
2. Communities to Follow
3. Companies to Keep an Eye On
4. Blogs to Read
5. Whitepapers
6. Great YouTube Channels
7. Great Podcasts
8. Newsletters
9. LinkedIn, Twitter, TikTok, and Instagram Influencers to Follow
10. Courses
11. Certifications
12. Conferences

Link: DataExpert-io/data-engineer-handbook

9. Data Engineering Wiki

The Data Engineering Wiki repository is a community-driven wiki that provides a comprehensive resource for learning data engineering. This repository covers a wide range of topics, including data pipelines, data warehousing, and data modeling.

Data Engineering Wiki includes:

1. Data Engineering Concepts
2. Frequently Asked Questions about Data Engineering
3. Guides on How to Make Data Engineering Decisions
4. Commonly Used Tools for Data Engineering
5. Step-by-Step Guides for Data Engineering Tasks
6. Learning Resources

Link: data-engineering-community/data-engineering-wiki

10. Data Engineering Practice

Data Engineering Practice offers a hands-on approach to learning data engineering. It provides practice projects and exercises to help you apply your knowledge and skills in real-world scenarios. By working through these projects, you will gain practical experience and build a portfolio that showcases your data engineering capabilities.

Data Engineering Practice Problems include exercises on:

1. Downloading Files
2. Web Scraping + Downloading + Pandas
3. Boto3 AWS + s3 + Python.
4. Convert JSON to CSV + Ragged Directories
5. Data Modeling for Postgres + Python
6. Ingestion and Aggregation with PySpark
7. Using Various PySpark Functions
8. Using DuckDB for Analytics and Transforms
9. Using Polars Lazy Computation

Link: danielbeach/data-engineering-practice

Final Words

Mastering data engineering requires dedication, persistence, and a passion for learning new concepts and tools. These 10 GitHub repositories provide a wealth of information and resources to help you become a professional data engineer and keep you updated on current trends.

Whether you are just starting or an experienced data engineer, I encourage you to explore these resources, contribute to open-source projects, and stay engaged with the vibrant data engineering community on GitHub.

Abid Ali Awan (@1abidaliawan) is a certified data scientist professional who loves building machine learning models. Currently, he is focusing on content creation and writing technical blogs on machine learning and data science technologies. Abid holds a Master's degree in technology management and a bachelor's degree in telecommunication engineering. His vision is to build an AI product using a graph neural network for students struggling with mental illness.

More On This Topic

• 10 GitHub Repositories to Master Machine Learning
• 10 GitHub Repositories to Master Computer Science
• 10 GitHub Repositories to Master MLOps
• 10 GitHub Repositories to Master Python
• Learn Data Engineering From These GitHub Repositories
• Learn Data Science From These GitHub Repositories

In a move to address the evolving needs of small and medium-sized enterprises (SMEs) and hosted IT service providers, AMD has introduced the AMD EPYC 4004 Series processors.

These new offerings, announced today, complement AMD’s existing EPYC server CPU lineup, providing cost-optimised solutions without compromising on performance and enterprise-class features.

Powered by the efficient “Zen 4” architecture, the AMD EPYC 4004 Series CPUs offer enterprise-grade performance, scalability, and modern security features, catering to price-conscious buyers.

Notably, a server equipped with a single AMD EPYC 4564P CPU outperforms an Intel Xeon E-2488 CPU by 1.8 times in terms of performance per dollar.

John Morris, corporate vice president of the Enterprise and HPC Business Group at AMD, emphasised the significance of these processors for businesses that historically had to settle for IT solutions that didn’t fully meet their requirements. “Based on the same technologies that power the most demanding data centres in the world, the AMD EPYC 4004 Series processors are offered at an optimised acquisition cost for customers in small and medium-sized businesses seeking to drive better business outcomes,” he said.

The AMD EPYC 4004 Series processors are engineered to deliver robust, general-purpose computing in a single-socket package, facilitating highly performant rack scale, multi-node, and tower configurations, particularly suitable for scenarios where system cost and infrastructure constraints are crucial considerations.

Key industry players have expressed their support for AMD’s initiative. Kamran Amini, Vice President and General Manager for Server, Storage & Software Defined Solutions at Lenovo, praised AMD’s efforts in expanding its EPYC processor roadmap to address a broader market segment with affordable yet high-performance capabilities.

OVHcloud’s chief product and technology officer, Yaniv Fdida, echoed the sentiment, expressing enthusiasm about adding AMD EPYC 4004 CPU-powered solutions to their Bare Metal portfolio, emphasising the potential for flexibility and performance-price ratio benefits in data centers.

Supermicro’s SVP Marketing and Network Security, Michael McNerney, highlighted the enhanced value brought by AMD EPYC 4004 Series CPUs to customers seeking cost-effective and easy-to-deploy solutions, particularly in workload performance optimization for hosting, content delivery, and cloud workloads.

Overall, the AMD EPYC 4004 CPU-powered servers promise a compelling balance of performance, scalability, and affordability, catering to a wide range of enterprise solutions. Supported by leading partners such as Altos, ASRock Rack, Gigabyte, MSI, New Egg, Tyan, and others, these processors signify AMD’s commitment to meeting the diverse needs of growing businesses.

The post AMD Unveils EPYC 4004 Processors to Compete with Intel’s Xeon Processors appeared first on AIM.

Google announced a new model, Gemini 1.5 Flash, at the Google I/O 2024. It’s a lightweight AI model optimised for speed and efficiency with a massive context window of 1M tokens.

Designed to handle tasks that require quick responses, it is capable of multimodal reasoning, which means it has the ability to simultaneously process and understand various types of data such as text, images, audio, and video.

It is a valuable tool for situations where time and efficiency are crucial and can be used in various applications from a customer service chatbot and generating captions or images for social media posts, to scientific research and business analytics.

People are still underestimating the value of Gemini 1.5 Flash.
For $0.35, you can get 1 million tokens and start building natively multi-modal projects.
The cost + latency + context window size + intelligence of Flash is going to create so many new startups.

— Logan Kilpatrick (@OfficialLoganK) May 18, 2024

“Gemini 1.5 Flash excels at summarisation, chat applications, image and video captioning, data extraction from long documents and tables, and more,” wrote Demis Hassabis, the CEO of Google DeepMind.

Hassabis further added that Google created Gemini 1.5 Flash to provide developers with a model that was lighter and less expensive than the Gemini 1.5 Pro version.

Despite being lighter in weight than Gemini Pro, Gemini 1.5 Flash is just as powerful. This is because it’s been trained through a process called “distillation”, where the most essential knowledge and skills from Gemini Pro are transferred to 1.5 Flash but in a way that makes the Flash model smaller and more efficient.

In addition to being the fastest in the Gemini family, it’s also more cost efficient to use, making it a faster and less expensive way for developers building their own AI products and services.

How Does Gemini 1.5 Flash Compare to Other Models?

Source: X

Many users tested Gemini 1.5 Flash and compared it with other models and in most cases 1.5 Flash performed impressively.

Over the weekend, I tested 3 LLMs to get relevancy score
1. Haiku
2. Gemini Flash 1.5
3. Perplexity: Llama3 Sonar 8B
Haiku didn't care to follow my instructions most of the time, btw I used claude to write the prompt.
Gemini Flash worked pretty well
Perplexity worked really…

— Praveen Kumar | Building MevinX (@PraveenInPublic) May 20, 2024

When compared with GPT-4o, a user posted that 1.5 Flash performed almost as well as GPT-4o on the StaticAnalysisEval benchmark. Additionally, it is faster and more cost-effective than GPT-4o, making it a compelling alternative.

A user tested GPT- 3.5 Turbo, Claude Haiku, and Gemini 1.5 Flash to check which model aligns most closely with GPT-4o in terms of accuracy for a specific classification task. Flash emerged as the clear winner.

Another posted that Gemini 1.5 Flash was better than Llama-3-70b on long context tasks. “It’s way faster than my locally hosted 70b model (on 4*A6000) and hallucinates less. The free of charge plan is good enough for me to do prompt engineering for prototyping,” he wrote.

A user ran 1.5 Flash on some evals for automatically triaging vulnerabilities in code, and did the same with GPT-4-Turbo hosted on Azure, Llama-3 70B hosted on Groq, and GPT-4o hosted on OpenAI as well.

“It’s very fast and very cheap. The results were pretty much on par with the other models in terms of accuracy,” he concluded.

I played with Google's new Gemini 1.5 Flash model over the weekend and was quite impressed.
It's not the best model out there, but can be very powerful if it works for your use case.
It's more verbose, but very fast and very cheap.
I ran it on some of our evals for… pic.twitter.com/mJIff2gerA

— Stefan Streichsbier (@s_streichsbier) May 20, 2024

Another user ran various tests for both Gemini Flash as well as GPT-4o and agreed that Google’s new model is impressive – cheaper, sometimes faster, and gives similar results to GPT-4o. “A combination of the two using LLM agentic workflow is the solution,” he added.

The new gemini-flash is 19x cheaper than gpt-4o & nearly as good.
But I don't trust benchmarks.
So I run my own tests:
test #1 → analyze youtube for me pic.twitter.com/E1bZsbjqzj

— Ruben Hassid (@RubenHssd) May 19, 2024

However, some have also raised concerns about the model’s low rate limit that is creating roadblocks in using it in production in any capacity.

Source: X

Interesting Use Cases of Gemini 1.5 Flash

Online users have been trying their hands on the model and are coming up with interesting use cases.

DIY-Astra, a multi-modal AI assistant powered by Gemini 1.5 Flash

Introducing DIY-Astra, a small but powerful web app powered by Gemini 1.5 Flash.
Astra will tell you anything it sees in the camera, essentially in real-time.
I was so impressed when I saw that it can also solve visual questions as well.
Repo in the comment. pic.twitter.com/eHUHPGEMHg

— Pietro Schirano (@skirano) May 16, 2024

The 1M token context, low cost, and high speed of Gemini 1.5 Flash make it a perfect tool to create exciting applications like these.

Gemini 1.5 Flash for WebScrapping

Gemini 1.5 Flash is ideal for web scraping. It simplifies the process by eliminating the need for HTML selectors and adapts to various HTML structures across devices, countries, and products. The model works efficiently with any web page technology, including JavaScript and pre-rendered HTML.

I'm testing #Gemini 1.5 Flash for #WebScrapping and the results are amazing
Gemini 1.5 Flash is a multimodal, lightweight, and affordable AI model (35 cents per million input tokens) for web scraping.
Here’s why AI is great for scraping:
No more dealing with HTML selectors.… pic.twitter.com/5wm2kCiUnp

— Xavi Ramirez (@xaviramirezcom) May 20, 2024

Analyse a Video to Produce Script

An online user gave Gemini 1.5 Flash a video recording of him shopping and it generated the Selenium code of the site in just about 5 seconds.

This is mind blowing
I gave Gemini 1.5 Flash video recording of me shopping and it gave me Selenium code in ~5 seconds.
This can change so many things. pic.twitter.com/Ojm6aueLe7

— Min Choi (@minchoi) May 18, 2024

Gemini-1.5-Flash as a Copilot in VSCode

By connecting CodeGPT with Google AI Studio, you can leverage the power of Gemini 1.5 Flash to enhance your coding experience.

Gemini-1.5-flash as a Copilot in VSCode is amazing!
You can now use this model by connecting CodeGPT with Google AI Studio.@codegptAI + @googleaistudio
In this video, I show how CodeGPT manages to get the entire context of the "Quick Fix" section and Gemini provides a… pic.twitter.com/E6eGczLgtb

— Daniel San (@dani_avila7) May 15, 2024

A Great Option for Voice AI

Gemini 1.5 Flash is a great option for voice AI, with first token around 500 ms and 150 tokens/s.

Gemini 1.5 Flash is a game changer for voice-based products. Adding it to Voqal really shows how easy it will be to interact with machines in the future.
It took <5min to "teach" my assistant how to watch my CI builds and alert me when they finish. Zero keywords. Zero wake… pic.twitter.com/t7NFJgRkxT

— Voqal (@voqaldev) May 16, 2024

Gemini YouTube Researcher

Let Gemini be your YouTube researcher. Simply input a topic, and the AI analyses relevant videos to deliver a comprehensive summary, simplifying your research by extracting key insights efficiently.

1. Gemini YouTube Researcher
– Listens to videos & delivers topical reports.
– Write a topic and AI will analyze relevant videos & provide a comprehensive report. pic.twitter.com/pQ0EkMPAXg

— Saumya Singh (@saumya1singh) May 20, 2024

This shows that with Gemini 1.5 Flash’s cost, latency, and 1M tokens context, alongside the OpenAI GPT-4o, which is also plausibly a lightweight model, the possibilities are endless.

The post With Google’s Gemini 1.5 Flash, the Possibilities are Endless appeared first on AIM.

Recently, Microsoft CEO Satya Nadella unveiled Windows’ latest feature: Recall. This isn’t just a keyword search; it’s a semantic search, delving deep into your digital history to recreate moments from the past.

“Recall isn’t just about documents,” Nadella explained in an interview with The Wall Street Journal. “It’s about reliving experiences, reimagining the past with clarity and detail.”

So, how does this work?

Windows PCs begin capturing screenshots of users’ activities, feeding this data into a sophisticated AI model embedded directly within the devices. Through neural processing, every image and interaction will become searchable, even extending to photographs.

Understandably, there are some concerns, with Elon Musk, in a characteristic tweet, saying, “This is a Black Mirror episode. Definitely turning this ‘feature’ off.”

A Good ‘Recall’

This comes in the backdrop of Microsoft Build, the company’s annual developer conference, which will begin today in Seattle. The company is set to showcase its latest AI projects following OpenAI and Google’s events hosted this month.

OpenAI announced the ChatGPT desktop app, powered by the latest GPT-4o model, which changed everything. This AI companion boasts real-time screen reading capabilities, positioning itself as the go-to colleague for assistance in times of need.

Earlier, OpenAI CEO Sam Altman had expressed views of AI systems acting like a “senior employee” who can engage with users much like a trusted employee would with a CEO. This includes the ability to push back, reason, and access emails within specified constraints.

Altman emphasised that an AI assistant shouldn’t be like an “agent” but like a “senior employee”. He pointed out that a senior employee would question requests that appear illogical, while an agent would unquestioningly follow commands.

In the meantime, Google introduced AI Teammate, powered by Gemini, which is aimed at streamlining workflow and communication within teams, minus screen recording.

This new feature will significantly reduce workloads by handling mundane tasks and participating in team communications, potentially transforming the employee from a helpful colleague to an overbearing overseer.

Forget Managers, the AI Employee is Here

As tech giants are announcing AI features for the workplace, in an interesting development, UK-based startup Artisan AI is on a mission to create the most advanced human-like digital workers, called Artisans.

So far, they have released Ava, a sales representative Artisan. Ava operates as a business development representative (BDR) who streamlines the entire outbound sales process, requiring only a brief 10-minute conversation for setup.

Ava not only formulates the user’s ideal customer profile (ICP) but also navigates a vast database of over 270 million contacts. From meticulous lead research to crafting and dispatching highly tailored email sequences, Ava executes each task with precision, seamlessly booking meetings into sales reps’ calendars.

A new era is upon us, with AI employees and humans working in symbiosis.
We've just released Ava, The Sales Rep Artisan. She's an AI BDR on steroids, and she's available to hire now.
The best part? Manage everything by chatting to her.
Learn more: https://t.co/Xq15n0B85J pic.twitter.com/nynI97wfwY

— Artisan (@GetArtisanAI) January 15, 2024

They have also recently launched Artisan Labs, their research lab, where they are pioneering the foundational technologies required to create human-like digital workers.

Time to be Friends with Your AI Colleague

In today’s rapidly evolving landscape, AI’s capacity to comprehend emotions emerges as a powerful asset rather than a cause for concern. Despite initial reservations regarding privacy, these AI systems can be instrumental in enhancing employee well-being.

By detecting signs of stress or excessive screen time, managers can intervene proactively, potentially mitigating the risk of burnout or mental health issues. Moreover, in the context of rising concerns about workplace suicides, AI monitoring could provide critical insights into identifying and addressing distressing behaviours among colleagues.

Moreover, humans naturally seek connections and understanding from others. In this context, AI’s ability to comprehend emotions can serve as a catalyst for fostering deeper relationships and facilitating a more supportive workplace culture.

Recent advancements in robotics and AI have fundamentally altered our perception of technology’s role in shaping social dynamics. By integrating emotional intelligence into AI systems, we not only enhance their efficacy but also redefine them as invaluable allies in promoting employee well-being and fostering positive workplace relationships.

The post Why is Microsoft Copilot Employees’ Worst Nightmare? appeared first on AIM.

At a special event ahead of Microsoft Build 2024, Microsoft announced its vision for a new era of AI PCs powered by Qualcomm’s Snapdragon X processors.

The new Surface Pro 10 and Surface Laptop 6 are the first Microsoft devices optimised for AI experiences. They are powered by Qualcomm’s Snapdragon X Elite and Plus processors, which deliver a massive 45 trillion operations per second (TOPS) of AI performance thanks to the integrated Neural Processing Unit (NPU).

This enables advanced on-device AI features like focus, auto-framing, background blur and other effects to device cameras and built-in microphones.

Microsoft is touting these new ARM-based Surface devices as the “most powerful Windows PCs ever built”. The company claims the Snapdragon X Elite chip makes the Surface Pro 10 up to 90% faster than the previous Intel-based Surface Pro 9, while providing a significantly longer battery life of up to 19 hours.

Recently, Apple also announced their M4 chips with their iPads which had 36 TOPS. But we are yet to see how Apple will optimise their M4 chips with their Macbooks, especially after the launch of the Surface laptops powered by Snapdragon X Elite chips.

Before the launch of Snapdragon X Elite chips, M4 were said to be “faster than the neural processing unit of any AI PC today”. However, a comparison between the iPad with Surface laptops won’t make any sense, especially with Apple about to unveil a new lineup of Macbooks powered by M4 chips on WWDC event.

It would be interesting to see how Apple will manage to reclaim their position as the fastest AI computer manufacturer.

The post Microsoft and Qualcomm Unleash 45 TOPS ARM Snapdragon X Elite Chips in Surface Laptops to Compete With Apple Silicon appeared first on AIM.

It appears that concerns about being ghosted by your interviewer are a thing of the past, all thanks to “Alex.”

But who exactly is Alex?

US-based startup Apriora AI has leveraged AI to streamline the hiring process. Their flagship product, Alex, represents a paradigm shift in recruitment methodology by seamlessly integrating advanced technology into the interview process.

Unlike traditional interviewers, Alex operates as a two-way AI interface, capable of conducting live video interviews with job candidates. This cutting-edge technology provides applicants with immediate feedback and a more transparent hiring experience.

3. Apriora (@aprioraai)
Round: $2.8M Seed
Apriora built an AI interviewer named Alex, which conducts live video and phone interviews.
Alex provides a personalized, conversational experience and generates comprehensive hiring insights.
Live interview: pic.twitter.com/TUHP5ff7Ff

— Chief AI Officer (@chiefaioffice) May 19, 2024

One of the notable features of Alex is its unparalleled capacity to manage interviews. Unlike humans, Alex does not require breaks or downtime, enabling it to conduct interviews continuously without interruption.

This perpetual availability ensures that companies can maintain a steady flow of interviews regardless of external factors such as weather conditions or staffing constraints.

Moreover, Alex’s ability to review resumes and conduct interviews non-stop throughout the day significantly increases the likelihood of job seekers securing interviews. By facilitating a higher volume of interviews, Apriora AI aims to enhance opportunities for both employers and applicants in the competitive job market.

Further, various platforms are now turning to AI models to streamline the hiring process. The platforms present recruiters and HR professionals with an array of AI tools to navigate this season of talent acquisition.

Some of the existing Indian generative AI platforms for recruiters and HR professionals include MachineHack for Enterprises, Oracle Recruiting, and Zoho Recruit.

The post Candidates Will Soon Be Interviewed by AI appeared first on AIM.

Intel has revealed that starting from the third quarter of 2024, its highly anticipated client processors, codenamed Lunar Lake, are slated to power over 80 fresh laptop designs across more than 20 original equipment manufacturers (OEMs).

These processors are primed to usher in a new era of AI performance on a global scale for Copilot+ PCs, set forward by Microsoft.

Underlining the significance of this development, Michelle Johnston Holthaus, Executive Vice President and General Manager of the Client Computing Group at Intel, emphasised the breakthrough power efficiency and compatibility of the x86 architecture. “With breakthrough power efficiency, the trusted compatibility of x86 architecture and the industry’s deepest catalogue of software enablement across the CPU, GPU and NPU, we will deliver the most competitive joint client hardware and software offering in our history with Lunar Lake and Copilot+,” he said.

An AI PC, comprising a CPU, GPU, and NPU, is tailored with specific AI acceleration capabilities. The NPU, in particular, serves as a specialised accelerator for AI and machine learning tasks directly on the PC, bypassing the need for cloud processing. The rising importance of AI PCs stems from the growing necessity to automate and optimise tasks on personal computers.

Lunar Lake is anticipated to revolutionise mobile processing for AI PCs, boasting over three times the AI performance compared to its predecessors. With an impressive 40+ NPU tera operations per second (TOPS), Intel’s next-gen processors are poised to deliver the capabilities required for the upcoming Copilot+ experiences. Moreover, Lunar Lake will feature over 60 GPU TOPS, amounting to more than 100 platform TOPS.

“The launch of Lunar Lake will bring meaningful fundamental improvements across security, battery life, and more thanks to our deep co-engineering partnership with Intel. We are excited to see Lunar Lake come to market with a 40+ TOPS NPU which will deliver Microsoft’s Copilot+ experiences at scale when available,” said Pavan Davuluri, Corporate Vice President of Windows + Devices at Microsoft.

Recognising the importance of both hardware innovation and software enablement, Intel is actively collaborating with over 100 independent software vendors through its AI PC Acceleration Program. This initiative aims to enhance AI PC experiences across various domains, including personal assistants, audio effects, content creation, gaming, security, streaming, and video collaboration.

According to reports, AMD is also coming up with its new APU, Ryzen 8050, featuring Zen 5 CPU and XDNA 2 NPU architecture for AI PC workloads. This will boast a performance of around 50 TOPS, ideal for running Microsoft’s AI PC goals.

The post Intel Lunar Lake Arriving Q3 2024 with 40+ TOPS for AI PCs appeared first on AIM.

Geoffrey Hinton, the godfather of AI, has made a pathbreaking dictum defying conventional wisdom – AI models can have sentiments. His memory of seeing an emotional robot in 1973 is even more impressive.

In a recent interview, Hinton spoke about witnessing an “emotional” robot in Edinburgh.

With its grippers in place, the robot could assemble a toy automobile if the parts were correctly arranged. But when the parts were dispersed, the robot behaved differently; it seemed “crossed” or irritated, much like a human would when faced with a challenging or unclear task.

This observation, dating back over four decades, underscores the potential for AI and robotics in exhibiting behaviours that we typically associate with human emotions. Hinton’s insights continue to push the boundaries of what we understand about consciousness and emotion in machines.

The Nature of AI Emotions

In a recent podcast, OpenAI chief Sam Altman predicted that the development of AI will force individuals to forge deeper human ties.

Altman states, “The broad category of new kinds of art, entertainment, is more akin to interpersonal relationships. I’m unsure whether we’ll arrive in five years, nor do I know the job title. However, I believe that amazing in-person human encounters will be prioritised.”

Altman had previously admitted that he was “a little bit scared” of ChatGPT. He advised, “We have to exercise caution here, and people ought to be relieved that we are a little afraid of this. If I said I wasn’t, you should not trust me or be sad that I work here.

Sundar Pichai, the CEO of Google, agreed with Altman. “We all refer to one part of this as the “black box” in the field. You can’t explain why it stated this or got it wrong, and you don’t fully comprehend,” he said.

Elon Musk, the owner of X (formerly Twitter), is also very outspoken about his worries around AI. He has called the progress of ChatGPT “concerning” on multiple occasions. In another post, he accused the chatbot of being “too woke”.

In a blog last month, Microsoft founder Bill Gates elaborated on the risk. “There’s a possibility that AIs will run out of control. Could a machine decide that humans are a threat, conclude that its interests differ from ours, or stop caring about us? Possibly, but this problem is no more urgent today than it was before the AI developments of the past few months.”

Understanding AI and Vice-versa

A week ago, CRED founder Kunal Shah questioned if readers would be able to tell a human-written news article from an AI-generated one. About 62% of users said they wouldn’t. However, research claims otherwise.

If news articles were written by AI and not humans would you be able to know?

— Kunal Shah (@kunalb11) May 19, 2024

According to experts, humans can detect the existence of another intelligent entity while engaging with chatbots and respond to it as a relevant source of engagement. The key topic is whether and how chatbots may have comparable levels and kinds of social influence on humans, even though research has generally acknowledged the social impact of chatbots.

For example, the fluidity with which ChatGPT spoke in the recent demonstration of GPT-4o, the recently announced premier large language model, blew the minds of many users.

It responded almost instantly, expressed a wide range of emotions, altered the volume and pacing of its speech, and could even sing.[

Perhaps even more remarkable was that it could hear. It could distinguish different breathing patterns, identify speakers by voice in a group conversation, harmonise with itself, and even respond to interruptions.

Similarly, at Google I/O, Google introduced AI Teammates and NotebookL, which help people in their work and teach them things quickly and meaningfully.

The boundaries between people and robots are getting fuzzier. As these technologies advance and become more integrated into our lives, we consider them independent social entities that can comprehend and respond to our wants and requirements.

This change affects our understanding of ourselves, relationships with others, and how we engage with technology.

Humans are inherently drawn to other people and their understanding. We seek connections with people who help us understand who we are and where we fit.

The actual, perceived, suggested, and attributed presence of others shapes human perceptions, behaviours, and experiences. Recent developments in robotics and AI have fundamentally changed how we perceive the origins and mechanisms of these social impacts.

A 2023 study suggests that artificial intelligence-generated faces have become indistinguishable from human faces. These indications help people feel more connected to AI and view it as more human. Moreover, AI frequently conveys gender cues and cultural preconceptions about human aid, which might endearingly and familiarly feel natural.

The post AI Models, Too, Have Feelings You Know appeared first on AIM.