Generative AI
Google announces Gemini: it’s most powerful AI model, powered by TPU chips
Google claims it has developed a new Generative Artificial Intelligence (GenAI) system and Large Language Model (LLM) more powerful than any currently on the market, including technology developed by ChatGPT creator OpenAI. Gemini can summarize text, create images and answer questions. Gemini was trained on Google’s Tensor Processing Units v4 and v5e.
Google’s Bard is a generative AI based on the PaLM large language mode. Starting today, Gemini will be used to give Bard “more advanced reasoning, planning, understanding and more,” according to a Google blog post.
While global users of Google Bard and the Pixel 8 Pro will be able to run Gemini now, an enterprise product, Gemini Pro, is coming on Dec. 13th. Developers can sign up now for an early preview in Android AICore.
Gemini comes in three model sizes: Ultra, Pro and Nano. Ultra is the most capable, Nano is the smallest and most efficient, and Pro sits in the middle for general tasks. The Nano version is what Google is using on the Pixel, while Bard gets Pro. Google says it plans to run “extensive trust and safety checks” before releasing Gemini Ultra to select groups.
Gemini can code in Python, Java, C++, Go and other popular programming languages. Google used Gemini to upgrade Google’s AI-powered code generation system, AlphaCode. Next, Google plans to bring Gemini to Ads, Chrome and Duet AI. In the future, Gemini will be used in Google Search as well.
……………………………………………………………………………………………………………………………………………………
Market Impact:
Gemini’s release and use will present a litmus test for Google’s technology following a push to move faster in developing and releasing AI products. It coincides with a period of turmoil at OpenAI that has sent tremors through the tight knit AI community, suggesting the industry’s leaders is far from settled.
The announcement of the new GenAI software is the latest attempt by Google to display its AI portfolio after the launch of ChatGPT about a year ago shook up the tech industry. Google wanted outside customers to perform testing on the most advanced version of Gemini before releasing it more widely, said Demis Hassabis, chief executive officer of Google DeepMind.
“We’ve been pushing forward with a lot of focus and intensity,” Hassabis said, adding that Gemini likely represented the company’s most ambitious combined science and engineering project to date.
Google said Wednesday it would offer a range of AI programs to customers under the Gemini umbrella. It touted the software’s ability to process various media, from audio to video, an important development as users turn to chatbots for a wider range of needs.
The most powerful Gemini Ultra version outperformed OpenAI’s technology, GPT-4, on a range of industry benchmarks, according to Google. That version is expected to become widely available for software developers early next year following testing with a select group of customers.
………………………………………………………………………………………………………………………………………………………………………………………………
Role of TPUs:
While most GenAI software and LLM’s are processed using NVIDIA’s neural network processors, Google’s tensor processing units (TPUs) will power Gemini. TPUs are custom-designed AI accelerators, which are optimized for training and inference of large AI models. Cloud TPUs are optimized for training large and complex deep learning models that feature many matrix calculations, for instance building large language models (LLMs). Cloud TPUs also have SparseCores, which are dataflow processors that accelerate models relying on embeddings found in recommendation models. Other use cases include healthcare, like protein folding modeling and drug discovery.
Google’s custom AI chips, known as tensor processing units, are embedded in compute servers at the company’s data center. Photo Credit: GOOGLE
…………………………………………………………………………………………………………………………………
Competitors:
Gemini and the products built with it, such as chatbots, will compete with OpenAI’s GPT-4, Microsoft’s Copilot (which is based on OpenAI’s GPT-4), Anthropic’s Claude AI, Meta’s Llama 2 and more. Google claims Gemini Ultra outperforms GPT-4 in several benchmarks, including the massive multitask language understanding general knowledge test and in Python code generation.
…………………………………………………………………………………………………………………………………
References:
Everything to know about Gemini, Google’s new AI model (blog.google)
Google Reveals Gemini, Its Much-Anticipated Large Language Model (techrepublic.com)
MTN Consulting: Generative AI hype grips telecom industry; telco CAPEX decreases while vendor revenue plummets
Ever since Generative (Gen) AI burst into the mainstream through public-facing platforms (e.g. ChatGPT) late last year, its promising capabilities have caught the attention of many. Not surprisingly, telecom industry execs are among the curious observers wanting to try Gen AI even as it continues to evolve at a rapid pace.
MTN Consulting says the telecom industry’s bond with AI is not new though. Many telcos have deployed conventional AI tools and applications in the past several years, but Gen AI presents opportunities for telcos to deliver significant incremental value over existing AI. A few large telcos have kickstarted their quest for Gen AI by focusing on “localization.” Through localization of processes using Gen AI, telcos vow to eliminate language barriers and improve customer engagement in their respective operating markets, especially where English as a spoken language is not dominant.
Telcos can harness the power of Gen AI across a wide range of different functions, but the two vital telco domains likely to witness transformative potential of Gen AI are networks and customer service. Both these domains are crucial: network demands are rising at an unprecedented pace with increased complexity, and delivering differentiated customer experiences remains an unrealized ambition for telcos.
Several Gen AI use cases are emerging within these two telco domains to address these challenges. In the network domain, these include topology optimization, network capacity planning, and predictive maintenance, for example. In the customer support domain, they include localized virtual assistants, personalized support, and contact center documentation.
Most of the use cases leveraging Gen AI applications involve dealing with sensitive data, be it network-related or customer-related. This will have major implications from the regulatory point of view, and regulatory concerns will constrain telcos’ Gen AI adoption and deployment strategies. The big challenge is the mosaic of complex and strict regulations prevalent in different markets that telcos will have to understand and adhere to when implementing Gen AI use cases in such markets. This is an area where third-party vendors will try to cash in by offering Gen AI solutions that are compliant with regulations in the respective markets.
Vendors will also play a key role for small- and medium-sized telcos in Gen AI implementation, by eliminating constraints due to the lack of technical expertise and HW/SW resources, skilled manpower, along with opex costs burden. Key vendors to watch out for in the Gen AI space are webscale providers who possess the ideal combination of providing cloud computing resources required to train large language models (LLM) coupled with their Gen AI expertise offered through pre-trained models.
Other key points from MTN Consulting on Gen AI in the telecom industry:
- Network operations and customer support will be key transformative areas.
- Telco workforce will become leaner but smarter in the Gen AI era.
- Strict regulations will be a major barrier for telcos.
- Vendors key to Gen AI integration; webscale providers set for more telco gains.
- Lock-in risks and rising software costs are key considerations in choosing vendors.
………………………………………………………………………………………………………………………………
Separately, MTN Consulting’s latest forecast called for $320B of telco capex in 2023, down only slightly from the $328B recorded in 2022. Early 3Q23 revenue reports from vendors selling into the telco market call this forecast into question. The dip in the Americas is worse than expected, and Asia’s expected 2023 growth has not materialized.
Key vendors are reporting significant YoY drops in revenue, pointing to inventory corrections, macroeconomic uncertainty (interest rates, in particular), and weaker telco spending. Network infrastructure sales to telcos (Telco NI) for key vendors Ericsson and Nokia dropped 11% and 16% YoY in 3Q23, respectively, measured in US dollars. By the same metric, NEC, Fujitsu and Samsung saw +1%, -52%, and -41% YoY growth; Adtran, Casa, and Juniper declined 29%, 7%, and 20%; fiber-centric vendors Clearfield, Corning, CommScope, and Prysmian all saw double digit declines.
MTN Consulting will update its operator forecast formally next month. In advance, this comment flags a weaker spending outlook than expected. Telco capex for 2023 is likely to come in around $300-$310B.
MTN Consulting’s Network Operator Forecast Through 2027: “Telecom is essentially a zero-growth industry”
MTN Consulting: Top Telco Network Infrastructure (equipment) vendors + revenue growth changes favor cloud service providers
Proposed solutions to high energy consumption of Generative AI LLMs: optimized hardware, new algorithms, green data centers
Cloud Service Providers struggle with Generative AI; Users face vendor lock-in; “The hype is here, the revenue is not”
Global Telco AI Alliance to progress generative AI for telcos
Amdocs and NVIDIA to Accelerate Adoption of Generative AI for $1.7 Trillion Telecom Industry
Bain & Co, McKinsey & Co, AWS suggest how telcos can use and adapt Generative AI
Generative AI Unicorns Rule the Startup Roost; OpenAI in the Spotlight
Generative AI in telecom; ChatGPT as a manager? ChatGPT vs Google Search
Generative AI could put telecom jobs in jeopardy; compelling AI in telecom use cases
MTN Consulting: Satellite network operators to focus on Direct-to-device (D2D), Internet of Things (IoT), and cloud-based services
MTN Consulting on Telco Network Infrastructure: Cisco, Samsung, and ZTE benefit (but only slightly)
MTN Consulting: : 4Q2021 review of Telco & Webscale Network Operators Capex
Proposed solutions to high energy consumption of Generative AI LLMs: optimized hardware, new algorithms, green data centers
Introduction:
Many generative AI tools rely on a type of natural-language processing called large language models (LLMs) to first learn and then make inferences about languages and linguistic structures (like code or legal-case prediction) used throughout the world. Some companies that use LLMs include: Anthropic (now collaborating with Amazon), Microsoft, OpenAI, Google, Amazon/AWS, Meta (FB), SAP, IQVIA. Here are some examples of LLMs: Google’s BERT, Amazon’s Bedrock, Falcon 40B, Meta’s Galactica, Open AI’s GPT-3 and GPT-4, Google’s LaMDA Hugging Face’s BLOOM Nvidia’s NeMO LLM.
The training process of the Large Language Models (LLMs) used in generative artificial intelligence (AI) is a cause for concern. LLMs can consume many terabytes of data and use over 1,000 megawatt-hours of electricity.
Alex de Vries is a Ph.D. candidate at VU Amsterdam and founder of the digital-sustainability blog Digiconomist published a report in Joule which predicts that current AI technology could be on track to annually consume as much electricity as the entire country of Ireland (29.3 terawatt-hours per year).
“As an already massive cloud market keeps on growing, the year-on-year growth rate almost inevitably declines,” John Dinsdale, chief analyst and managing director at Synergy, told CRN via email. “But we are now starting to see a stabilization of growth rates, as cloud provider investments in generative AI technology help to further boost enterprise spending on cloud services.”
Hardware vs Algorithmic Solutions to Reduce Energy Consumption:
Roberto Verdecchia is an assistant professor at the University of Florence and the first author of a paper published on developing green AI solutions. He says that de Vries’s predictions may even be conservative when it comes to the true cost of AI, especially when considering the non-standardized regulation surrounding this technology. AI’s energy problem has historically been approached through optimizing hardware, says Verdecchia. However, continuing to make microelectronics smaller and more efficient is becoming “physically impossible,” he added.
In his paper, published in the journal WIREs Data Mining and Knowledge Discovery, Verdecchia and colleagues highlight several algorithmic approaches that experts are taking instead. These include improving data-collection and processing techniques, choosing more-efficient libraries, and improving the efficiency of training algorithms. “The solutions report impressive energy savings, often at a negligible or even null deterioration of the AI algorithms’ precision,” Verdecchia says.
……………………………………………………………………………………………………………………………………………………………………………………………………………………
Another Solution – Data Centers Powered by Alternative Energy Sources:
The immense amount of energy needed to power these LLMs, like the one behind ChatGPT, is creating a new market for data centers that run on alternative energy sources like geothermal, nuclear and flared gas, a byproduct of oil production. Supply of electricity, which currently powers the vast majority of data centers, is already strained from existing demands on the country’s electric grids. AI could consume up to 3.5% of the world’s electricity by 2030, according to an estimate from IT research and consulting firm Gartner.
Amazon, Microsoft, and Google were among the first to explore wind and solar-powered data centers for their cloud businesses, and are now among the companies exploring new ways to power the next wave of AI-related computing. But experts warn that given their high risk, cost, and difficulty scaling, many nontraditional sources aren’t capable of solving near-term power shortages.
Exafunction, maker of the Codeium generative AI-based coding assistant, sought out energy startup Crusoe Energy Systems for training its large-language models because it offered better prices and availability of graphics processing units, the advanced AI chips primarily produced by Nvidia, said the startup’s chief executive, Varun Mohan.
AI startups are typically looking for five to 25 megawatts of data center power, or as much as they can get in the near term, according to Pat Lynch, executive managing director for commercial real-estate services firm CBRE’s data center business. Crusoe will have about 200 megawatts by year’s end, Lochmiller said. Training one AI model like OpenAI’s GPT-3 can use up to 10 gigawatt-hours, roughly equivalent to the amount of electricity 1,000 U.S. homes use in a year, University of Washington research estimates.
Major cloud providers capable of providing multiple gigawatts of power are also continuing to invest in renewable and alternative energy sources to power their data centers, and use less water to cool them down. By some estimates, data centers account for 1% to 3% of global electricity use.
An Amazon Web Services spokesperson said the scale of its massive data centers means it can make better use of resources and be more efficient than smaller, privately operated data centers. Amazon says it has been the world’s largest corporate buyer of renewable energy for the past three years.
Jen Bennett, a Google Cloud leader in technology strategy for sustainability, said the cloud giant is exploring “advanced nuclear” energy and has partnered with Fervo Energy, a startup beginning to offer geothermal power for Google’s Nevada data center. Geothermal, which taps heat under the earth’s surface, is available around the clock and not dependent on weather, but comes with high risk and cost.
“Similar to what we did in the early days of wind and solar, where we did these large power purchase agreements to guarantee the tenure and to drive costs down, we think we can do the same with some of the newer energy sources,” Bennett said.
References:
https://aws.amazon.com/what-is/large-language-model/
https://spectrum.ieee.org/ai-energy-consumption
https://www.crn.com/news/cloud/microsoft-aws-google-cloud-market-share-q3-2023-results/6
Amdocs and NVIDIA to Accelerate Adoption of Generative AI for $1.7 Trillion Telecom Industry
SK Telecom and Deutsche Telekom to Jointly Develop Telco-specific Large Language Models (LLMs)
AI Frenzy Backgrounder; Review of AI Products and Services from Nvidia, Microsoft, Amazon, Google and Meta; Conclusions
Amdocs and NVIDIA to Accelerate Adoption of Generative AI for $1.7 Trillion Telecom Industry
Amdocs and NVIDIA today announced they are collaborating to optimize large language models (LLMs) to speed adoption of generative AI applications and services across the $1.7 trillion telecommunications and media industries.(1)
Amdocs and NVIDIA will customize enterprise-grade LLMs running on NVIDIA accelerated computing as part of the Amdocs amAIz framework. The collaboration will empower communications service providers to efficiently deploy generative AI use cases across their businesses, from customer experiences to network provisioning.
Amdocs will use NVIDIA DGX Cloud AI supercomputing and NVIDIA AI Enterprise software to support flexible adoption strategies and help ensure service providers can simply and safely use generative AI applications.
Aligned with the Amdocs strategy of advancing generative AI use cases across the industry, the collaboration with NVIDIA builds on the previously announced Amdocs-Microsoft partnership. Service providers and media companies can adopt these applications in secure and trusted environments, including on premises and in the cloud.
With these new capabilities — including the NVIDIA NeMo framework for custom LLM development and guardrail features — service providers can benefit from enhanced performance, optimized resource utilization and flexible scalability to support emerging and future needs.
“NVIDIA and Amdocs are partnering to bring a unique platform and unmatched value proposition to customers,” said Shuky Sheffer, Amdocs Management Limited president and CEO. “By combining NVIDIA’s cutting-edge AI infrastructure, software and ecosystem and Amdocs’ industry-first amAlz AI framework, we believe that we have an unmatched offering that is both future-ready and value-additive for our customers.”
“Across a broad range of industries, enterprises are looking for the fastest, safest path to apply generative AI to boost productivity,” said Jensen Huang, founder and CEO of NVIDIA. “Our collaboration with Amdocs will help telco service providers automate personalized assistants, service ticket routing and other use cases for their billions of customers, and help the telcos analyze and optimize their operations.”
Amdocs counts more than 350 of the world’s leading telecom and media companies as customers, including 27 of the world’s top 30 service providers.(2) With more than 1.7 billion daily digital journeys, Amdocs platforms impact more than 3 billion people around the world.
NVIDIA and Amdocs are exploring a number of generative AI use cases to simplify and improve operations by providing secure, cost-effective and high-performance generative AI capabilities.
Initial use cases span customer care, including accelerating customer inquiry resolution by drawing information from across company data. On the network operations side, the companies are exploring how to proactively generate solutions that aid configuration, coverage or performance issues as they arise.
(1) Source: IDC, OMDIA, Factset analyses of Telecom 2022-2023 revenue.
(2) Source: OMDIA 2022 revenue estimates, excludes China.
Editor’s Note:
- Language models: These models, like OpenAI’s GPT-3, generate human-like text. One of the most popular examples of language-based generative models are called large language models (LLMs).
- Large language models are being leveraged for a wide variety of tasks, including essay generation, code development, translation, and even understanding genetic sequences.
- Generative adversarial networks (GANs): These models use two neural networks, a generator, and a discriminator.
- Unimodal models: These models only accept one data input format.
- Multimodal models: These models accept multiple types of inputs and prompts. For example, GPT-4 can accept both text and images as inputs.
- Variational autoencoders (VAEs): These deep learning architectures are frequently used to build generative AI models.
- Foundation models: These models generate output from one or more inputs (prompts) in the form of human language instructions.
https://www.nvidia.com/en-us/glossary/data-science/generative-ai/
https://blogs.nvidia.com/blog/2023/01/26/what-are-large-language-models-used-for/
Cloud Service Providers struggle with Generative AI; Users face vendor lock-in; “The hype is here, the revenue is not”
Global Telco AI Alliance to progress generative AI for telcos
Bain & Co, McKinsey & Co, AWS suggest how telcos can use and adapt Generative AI
Generative AI Unicorns Rule the Startup Roost; OpenAI in the Spotlight
Generative AI in telecom; ChatGPT as a manager? ChatGPT vs Google Search
Generative AI could put telecom jobs in jeopardy; compelling AI in telecom use cases
SK Telecom and Deutsche Telekom to Jointly Develop Telco-specific Large Language Models (LLMs)
SK Telecom and Deutsche Telekom announced that they signed a Letter of Intent (LOI) to jointly develop a telco-specific Large Language Models (LLMs) that will enable global telecommunication companies (telcos) to develop generative AI models easily and quickly. The LOI signing ceremony took place at SK Seorin Building located in Seoul with the attendance of key executives from both companies including Ryu Young-sang, CEO of SKT, Chung Suk-geun, Chief AI Global Officer of SKT, Tim Höttges, CEO of Deutsche Telekom, Claudia Nemat, Board Member Technology and Innovation of Deutsche Telekome, and Jonathan Abrahamson, Chief Product and Digital Officer of Deutsche Telekom.
SK Telecom and Deutsche Telekom to Jointly Develop Telco-specific LLM
This marks the first fruition of discussions held by the Global Telco AI Alliance, which was launched by SKT, Deutsche Telekom, E&, and Singtel, in July 2023, and lays the foundation to enter the global market. SKT and Deutsche Telekom plan to collaborate with AI companies such as Anthropic (Claude 2) and Meta (Llama2) to co-develop a multilingual – i.e, German, English, Korean, etc. – large language model (LLM) tailored to the needs of telcos. They plan to unveil the first version of the telco-specific LLM in the first quarter of 2024.
The telco-specific LLM will have a higher understanding of telecommunication service-related areas and customer’s intentions than general LLMs, making it suitable for customer services like AI contact center. The goal is to support telcos across the world, including Europe, Asia, and the Middle East, to develop generative AI services such as AI agents flexibly according to their respective environment. That will enable telcos to save both time and cost for developing large platforms, and secure new business opportunities and growth engines through AI innovation that shifts the paradigm in the traditional telecommunications industry. To this end, SKT and Deutsche Telekom plan to jointly develop AI platform technologies that telcos can use to create generative AI services to reduce both development time and cost.
For instance, when a telco tries to build an AI contact center based on generative AI, it itself will be able to build one that suits their environment more quickly and flexibly. In addition, AI can be applied to other areas such as network monitoring and on-site operations to increase efficiency, resulting in cost savings in the mid- to long-term.
Through this collaboration, the two companies will proactively respond to the recent surge in AI demand from telcos, while also promoting the expansion of the global AI ecosystem through the successful introduction of generative AI optimized for specific industries or domains.
“AI shows impressive potential to significantly enhance human problem-solving capabilities. To maximize its use especially in customer service, we need to adapt existing large language models and train them with our unique data. This will elevate our generative AI tools,” says Claudia Nemat, Member of the Board of Management for Technology and Innovation at Deutsche Telekom.
“Through our partnership with Deutsche Telekom, we have secured a strong opportunity and momentum to gain global AI leadership and drive new growth,” said Ryu Young-sang, CEO of SKT. “By combining the strengths and capabilities of the two companies in AI technology, platform and infrastructure, we expect to empower enterprises in many different industries to deliver new and higher value to their customers.”
References:
Global Telco AI Alliance to progress generative AI for telcos
Cloud Service Providers struggle with Generative AI; Users face vendor lock-in; “The hype is here, the revenue is not”
Bain & Co, McKinsey & Co, AWS suggest how telcos can use and adapt Generative AI
Generative AI Unicorns Rule the Startup Roost; OpenAI in the Spotlight
Generative AI in telecom; ChatGPT as a manager? ChatGPT vs Google Search
Generative AI could put telecom jobs in jeopardy; compelling AI in telecom use cases
Google Cloud infrastructure enhancements: AI accelerator, cross-cloud network and distributed cloud
Cloud infrastructure services market grows; AI will be a major driver of future cloud service provider investments
TPG, Ericsson launch AI-powered analytics, troubleshooting service for 4G/5G Mobile, FWA, and IoT subscribers
Cloud Service Providers struggle with Generative AI; Users face vendor lock-in; “The hype is here, the revenue is not”
Everyone agrees that Generative AI has great promise and potential. Martin Casado of Andreessen Horowitz recently wrote in the Wall Street Journal that the technology has “finally become transformative:”
“Generative AI can bring real economic benefits to large industries with established and expensive workloads. Large language models could save costs by performing tasks such as summarizing discovery documents without replacing attorneys, to take one example. And there are plenty of similar jobs spread across fields like medicine, computer programming, design and entertainment….. This all means opportunity for the new class of generative AI startups to evolve along with users, while incumbents focus on applying the technology to their existing cash-cow business lines.”
A new investment wave caused by generative AI is starting to loom among cloud service providers, raising questions about whether Big Tech’s spending cutbacks and layoffs will prove to be short lived. Pressed to say when they would see a revenue lift from AI, the big U.S. cloud companies (Microsoft, Alphabet/Google, Meta/FB and Amazon) all referred to existing services that rely heavily on investments made in the past. These range from the AWS’s machine learning services for cloud customers to AI-enhanced tools that Google and Meta offer to their advertising customers.
Microsoft offered only a cautious prediction of when AI would result in higher revenue. Amy Hood, chief financial officer, told investors during an earnings call last week that the revenue impact would be “gradual,” as the features are launched and start to catch on with customers. The caution failed to match high expectations ahead of the company’s earnings, wiping 7% off its stock price (MSFT ticker symbol) over the following week.
When it comes to the newer generative AI wave, predictions were few and far between. Amazon CEO Andy Jassy said on Thursday that the technology was in its “very early stages” and that the industry was only “a few steps into a marathon”. Many customers of Amazon’s cloud arm, AWS, see the technology as transformative, Jassy noted that “most companies are still figuring out how they want to approach it, they are figuring out how to train models.” He insisted that every part of Amazon’s business was working on generative AI initiatives and the technology was “going to be at the heart of what we do.”
There are a number of large language models that power generative AI, and many of the AI companies that make them have forged partnerships with big cloud service providers. As business technology leaders make their picks among them, they are weighing the risks and benefits of using one cloud provider’s AI ecosystem. They say it is an important decision that could have long-term consequences, including how much they spend and whether they are willing to sink deeper into one cloud provider’s set of software, tools, and services.
To date, AI large language model makers like OpenAI, Anthropic, and Cohere have led the charge in developing proprietary large language models that companies are using to boost efficiency in areas like accounting and writing code, or adding to their own products with tools like custom chatbots. Partnerships between model makers and major cloud companies include OpenAI and Microsoft Azure, Anthropic and Cohere with Google Cloud, and the machine-learning startup Hugging Face with Amazon Web Services. Databricks, a data storage and management company, agreed to buy the generative AI startup MosaicML in June.
If a company chooses a single AI ecosystem, it could risk “vendor lock-in” within that provider’s platform and set of services, said Ram Chakravarti, chief technology officer of Houston-based BMC Software. This paradigm is a recurring one, where a business’s IT system, software and data all sit within one digital platform, and it could become more pronounced as companies look for help in using generative AI. Companies say the problem with vendor lock-in, especially among cloud providers, is that they have difficulty moving their data to other platforms, lose negotiating power with other vendors, and must rely on one provider to keep its services online and secure.
Cloud providers, partly in response to complaints of lock-in, now offer tools to help customers move data between their own and competitors’ platforms. Businesses have increasingly signed up with more than one cloud provider to reduce their reliance on any single vendor. That is the strategy companies could end up taking with generative AI, where by using a “multiple generative AI approach,” they can avoid getting too entrenched in a particular platform. To be sure, many chief information officers have said they willingly accept such risks for the convenience, and potentially lower cost, of working with a single technology vendor or cloud provider.
A significant challenge in incorporating generative AI is that the technology is changing so quickly, analysts have said, forcing CIOs to not only keep up with the pace of innovation, but also sift through potential data privacy and cybersecurity risks.
A company using its cloud provider’s premade tools and services, plus guardrails for protecting company data and reducing inaccurate outputs, can more quickly implement generative AI off-the-shelf, said Adnan Masood, chief AI architect at digital technology and IT services firm UST. “It has privacy, it has security, it has all the compliance elements in there. At that point, people don’t really have to worry so much about the logistics of things, but rather are focused on utilizing the model.”
For other companies, it is a conservative approach to use generative AI with a large cloud platform they already trust to hold sensitive company data, said Jon Turow, a partner at Madrona Venture Group. “It’s a very natural start to a conversation to say, ‘Hey, would you also like to apply AI inside my four walls?’”
End Quotes:
“Right now, the evidence is a little bit scarce about what the effect on revenue will be across the tech industry,” said James Tierney of Alliance Bernstein.
Brent Thill, an analyst at Jefferies, summed up the mood among investors: “The hype is here, the revenue is not. Behind the scenes, the whole industry is scrambling to figure out the business model [for generative AI]: how are we going to price it? How are we going to sell it?”
………………………………………………………………………………………………………………
References:
https://www.ft.com/content/56706c31-e760-44e1-a507-2c8175a170e8
https://www.wsj.com/articles/companies-weigh-growing-power-of-cloud-providers-amid-ai-boom-478c454a
https://www.techtarget.com/searchenterpriseai/definition/generative-AI?Offer=abt_pubpro_AI-Insider
Global Telco AI Alliance to progress generative AI for telcos
Curmudgeon/Sperandeo: Impact of Generative AI on Jobs and Workers
Bain & Co, McKinsey & Co, AWS suggest how telcos can use and adapt Generative AI
Generative AI Unicorns Rule the Startup Roost; OpenAI in the Spotlight
Generative AI in telecom; ChatGPT as a manager? ChatGPT vs Google Search
Generative AI could put telecom jobs in jeopardy; compelling AI in telecom use cases
Qualcomm CEO: AI will become pervasive, at the edge, and run on Snapdragon SoC devices
Global Telco AI Alliance to progress generative AI for telcos
- Four major global telcos joined forces to launch the Global Telco AI Alliance to accelerate AI transformation of the existing telco business and create new business opportunities with AI services.
- They signed a Multilateral MOU for cooperation in the AI business, which includes the co-development of the Telco AI Platform.
Deutsche Telekom, e&, Singtel and SK Telecom have established a new industry group that aims to progress generative AI. Called the Global Telco AI Alliance, it represents a coordinated effort by these four operators to accelerate the AI-fuelled transformation of their businesses, and to develop new, AI-powered business models.
The Telco AI Platform will serve as the foundation both for new services – like chatbots and apps – as well as enhancements to existing telco services. The alliance members plan to establish a working group whose task will be to hammer out co-investment opportunities and the co-development of said platform.
Members will also support one another in operating AI services and apps in their respective markets, and cooperate to foster the growth of a telco AI-based ecosystem.
As of today, all the operators have done is sign a memorandum of understanding (MoU), under which they pledge to carry out all this work. A signing ceremony took place in Seoul, Korea, and was attended – either in person or virtually – by the CEOs of e&, Singtel and SK Telecom, and Deutsche Telekom’s board member for technology and innovation, Claudia Nemat. The Global Telco AI Alliance will also have to ensure that any AI-based services they develop are capable of accounting for cultural differences. They won’t get very far if their virtual assistants make culturally insensitive recommendations, for example.
The seniority of these signatories represents a strong statement of intent though, and the group said it will discuss appointing C-level representatives from each member to the Alliance.
“In order to make the most of the possibilities of generative AI for our customers and our industry, we want to develop industry-specific applications in the Telco AI Alliance. I am particularly pleased that this alliance also stands for bridging the gap between Europe and Asia and that we are jointly pursuing an open-vendor approach. Depending on the application, we can use the best technology. The founding of this alliance is an important milestone for our industry,” said Claudia Nemat, Board Member Technology and Innovation at Deutsche Telekom.
“We recognize AI’s immense potential in reshaping the telecommunications landscape and beyond and are excited to embark on this transformative journey with the formation of the Global Telco AI Alliance. The alliance signifies a strategic commitment to driving innovation and fostering collaborative efforts. Our shared goal is to redefine industry paradigms, establish new growth drivers through AI-powered business models, and pave the way for a new era of strategic cooperation, guiding our industry towards an exciting and prosperous future,” said Khalifa Al Shamsi, CEO of e& life.
“This alliance will enable us and our ecosystem of partners to significantly expedite the development of new and innovative AI services that can bring tremendous benefits to both businesses and consumers. With our advanced 5G network, we are well-placed to leverage AI to ideate and co-create and are already using it to enhance our own customer service and employee experience, increase productivity and drive learning,” said Yuen Kuan Moon, Group Chief Executive Officer of Singtel.
It is not clear at this stage of proceedings whether the operators plan to develop their own in-house AI assets, or license them from the likes of OpenAI’s ChatGPT, or Google Bard. On the one hand, going with a third party that has done most of the legwork offers efficiencies, but on the other hand, the Global Telco AI Alliance might prefer an AI that specialises in telecoms, rather than a generalist.
Japanese vendor NEC showed earlier this month – with the launch of its own large language model (LLM) for enterprises in its home market – that generative AI isn’t necessarily the preserve of Silicon Valley big tech. It also highlighted the desire to develop localised AI for different languages.
The announcement also doesn’t attempt to grapple with any potential ethical pitfalls that might befall the Alliance. While it’s a fairly safe bet that responsible AI development will be an important consideration, it’s always better when companies make that clear.
Even big tech has come round to that way of thinking, with the launch earlier this week of the Frontier Model Forum. Established by Google, Microsoft, OpenAI and self-styled ethical AI company Anthropic, the group aims to advance the development of responsible artificial intelligence for the benefit of humanity.
References:
https://telecoms.com/522891/telcos-team-up-for-ai-platform-project/
https://telecoms.com/522865/google-microsoft-anthropic-and-openai-launch-ai-safety-body/
https://telecoms.com/522603/nec-launches-its-own-generative-ai/
Bain & Co, McKinsey & Co, AWS suggest how telcos can use and adapt Generative AI
Generative Artificial Intelligence (AI) uncertainty is especially challenging for the telecommunications industry which has a history of very slow adaptation to change and thus faces lots of pressure to adopt generative AI in their services and infrastructure. Indeed, Deutsche Telekom stated that AI poses massive challenges for telecom industry in this IEEE Techblog post.
Consulting firm Bain & Co. highlighted that inertia in a recent report titled, “Telcos, Stop Debating Generative AI and Just Get Going” Three partners stated network operators need to act fast in order to jump on this opportunity. “Speedy action trumps perfect planning here,” Herbert Blum, Jeff Katzin and Velu Sinha wrote in the brief. “It’s more important for telcos to quickly launch an initial set of generative AI applications that fit the company’s strategy, and do so in a responsible way – or risk missing a window of opportunity in this fast-evolving sector.”
Generative AI use cases can be divided into phases based on ease of implementation, inherent risk, and value:
………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………….
Telcos can pursue generative AI applications across business functions, starting with knowledge management:
Separately, a McKinsey & Co. report opined that AI has highlighted business leader priorities. The consulting firm cited organizations that have top executives championing an organization’s AI initiatives, including the need to fund those programs. This is counter to organizations that lack a clear directive on their AI plans, which results in wasted spending and stalled development. “Reaching this state of AI maturity is no easy task, but it is certainly within the reach of telcos,” the firm noted. “Indeed, with all the pressures they face, embracing large-scale deployment of AI and transitioning to being AI-native organizations could be key to driving growth and renewal. Telcos that are starting to recognize this is non-negotiable are scaling AI investments as the business impact generated by the technology materializes.”
…………………………………………………………………………………………………………………………………………………………………………………………………………………………………………….
Ishwar Parulkar, chief technologist for the telco industry at AWS, touted several areas that should be of generative AI interest to telecom operators. The first few were common ones tied to improving the customer experience. This includes building on machine learning (ML) to help improve that interaction and potentially reduce customer churn.
“We have worked with some leading customers and implemented this in production where they can take customer voice calls, translate that to text, do sentiment analysis on it … and then feed that into reducing customer churn,” Parulkar said. “That goes up another notch with generative AI, where you can have chat bots and more interactive types of interfaces for customers as well as for customer care agent systems in a call. So that just goes up another notch of generative AI.”
The next step is using generative AI to help operators bolster their business operations and systems. This is for things like revenue assurance and finding revenue leakage, items that Parulkar noted were in a “more established space in terms of what machine learning can do.”
However, Parulkar said the bigger opportunity is around helping operators better design and manage network operations. This is an area that remains the most immature, but one that Parulkar is “most excited about.” This can begin from the planning and installation phase, with an example of helping technicians when they are installing physical equipment.
“In installation of network equipment today, you have technicians who go through manuals and have procedures to install routers and base stations and connect links and fibers,” Parulkar said. “That all can be now made interactive [using] chat bot, natural language kind of framework. You can have a lot of this documentation, training data that can train foundational models that can create that type of an interface, improves productivity, makes it easier to target specific problems very quickly in terms of what you want to deploy.”
This can also help with network configuration by using large datasets to help automatically generate configurations. This could include the ability to help configure routers, VPNs and MPLS circuits to support network performance.
The final area of support could be in the running of those networks once they are deployed. Parulkar cited functions like troubleshooting failures that can be supported by a generative AI model.
“There are recipes that operators go through to troubleshoot and triage failure,” Parulkar said “A lot of times it’s trial-and-error method that can be significantly improved in a more interactive, natural language, prompt-based system that guides you through troubleshooting and operating the network.”
This model could be especially compelling for operators as they integrate more routers to support disaggregated 5G network models for mobile edge computing (MEC), private networks and the use of millimeter-wave (mmWave) spectrum bands.
Federal Communications Commission (FCC) Chairwoman Jessica Rosenworcel this week also hinted at the ability for AI to help manage spectrum resources.
“For decades we have licensed large slices of our airwaves and come up with unlicensed policies for joint use in others,” Rosenworcel said during a speech at this week’s FCC and National Science Foundation Joint Workshop. “But this scheme is not truly dynamic. And as demands on our airwaves grow – as we move from a world of mobile phones to billions of devices in the internet of things (IoT)– we can take newfound cognitive abilities and teach our wireless devices to manage transmissions on their own. Smarter radios using AI can work with each other without a central authority dictating the best of use of spectrum in every environment. If that sounds far off, it’s not. Consider that a large wireless provider’s network can generate several million performance measurements every minute. And consider the insights that machine learning can provide to better understand network usage and support greater spectrum efficiency.”
While generative AI does have potential, Parulkar also left open the door for what he termed “traditional AI” and which he described as “supervised and unsupervised learning.”
“Those techniques still work for a lot of the parts in the network and we see a combination of these two,” Parulkar said. “For example, you might use anomaly detection for getting some insights into the things to look at and then followed by a generative AI system that will then give an output in a very interactive format and we see that in some of the use cases as well. I think this is a big area for telcos to explore and we’re having active conversations with multiple telcos and network vendors.”
Parulkar’s comments come as AWS has been busy updating its generative AI platforms. One of the most recent was the launch of its $100 million Generative AI Innovation Center, which is targeted at helping guide businesses through the process of developing, building and deploying generative AI tools.
“Generative AI is one of those technological shifts that we are in the early stages of that will impact all organizations across the globe in some form of fashion,” Sri Elaprolu, senior leader of generative AI at AWS, told SDxCentral. “We have the goal of helping as many customers as we can, and as we need to, in accelerating their journey with generative AI.”
References:
https://www.bain.com/insights/telcos-stop-debating-generative-ai-and-just-get-going/
Deutsche Telekom exec: AI poses massive challenges for telecom industry
Generative AI in telecom; ChatGPT as a manager? ChatGPT vs Google Search
Generative AI could put telecom jobs in jeopardy; compelling AI in telecom use cases
Generative AI Unicorns Rule the Startup Roost; OpenAI in the Spotlight
Forbes: Cloud is a huge challenge for enterprise networks; AI adds complexity
Qualcomm CEO: AI will become pervasive, at the edge, and run on Snapdragon SoC devices
Bloomberg: China Lures Billionaires Into Race to Catch U.S. in AI
Bloomberg: China Lures Billionaires Into Race to Catch U.S. in AI
China’s tech sector has a new obsession: competing with U.S. titans like Google and Microsoft Corp. in the breakneck global artificial intelligence race. A ChatGPT-inspired global wave of AI activity is only just beginning in the next battle for supremacy in technology.
Billionaire entrepreneurs, mid-level engineers and veterans of foreign firms alike now harbor a remarkably consistent ambition: to outdo China’s geopolitical rival in a technology that may determine the global power stakes. Among them is internet mogul Wang Xiaochuan, who entered the field after OpenAI’s ChatGPT debuted to a social media firestorm in November. He joins the ranks of Chinese scientists, programmers and financiers — including former employees of ByteDance Ltd., e-commerce platform JD.com Inc. and Google — expected to propel some $15 billion of spending on AI technology this year.
For Wang, who founded the search engine Sogou that Tencent Holdings Ltd. bought out in a $3.5 billion deal less than two years ago, the opportunity came fast. By April, the computer science graduate had already set up his own startup and secured $50 million in seed capital. He reached out to former subordinates at Sogou, many of whom he convinced to come on board. By June, his firm had launched an open-source large language model and it’s already in use by researchers at China’s two most prominent universities.
“We all heard the sound of the starter pistol in the race. Tech companies, big or small, are all on the same starting line,” Wang, who named his startup Baichuan or “A Hundred Rivers,” told Bloomberg News. “China is still three years behind the US, but we may not need three years to catch up.”
The top-flight Chinese talent and financing flowing into AI mirrors a wave of activity convulsing Silicon Valley, which has deep implications for Beijing’s escalating conflict with Washington. Analysts and executives believe AI will shape the technology leaders of the future, much like the internet and smartphone created a corps of global titans. Moreover, it could propel applications from supercomputing to military prowess — potentially tilting the geopolitical balance.
China is a vastly different landscape — one reined in by US tech sanctions, regulators’ data and censorship demands, and Western distrust that limits the international expansion of its national champions. All that will make it harder to play catch-up with the US.
AI investments in the US dwarf that of China, totaling $26.6 billion in the year to mid-June versus China’s $4 billion, according to previously unreported data collated by consultancy Preqin.
China’s Catch-Up Game
The aggregate size of US deals in AI still outpaces China’s in 2023
Source: Preqin
Note: 2023 data up to June 14th
Yet that gap is already gradually narrowing, at least in terms of deal flow. The number of Chinese venture deals in AI comprised more than two-thirds of the US total of about 447 in the year to mid-June, versus about 50% over the previous two years. China-based AI venture deals also outpaced consumer tech in 2022 and early 2023, according to Preqin.
All this is not lost on Beijing. Xi Jinping’s administration realizes that AI, much like semiconductors, will be critical to maintaining China’s ascendancy and is likely to mobilize the nation’s resources to drive advances. While startup investment cratered during the years Beijing went after tech giants and “reckless expansion of capital,” the feeling is the Party encourages AI exploration.
It’s a familiar challenge for Chinese tech players. During the mobile era, a generation of startups led by Tencent, Alibaba Group Holding Ltd. and TikTok-owner ByteDance built an industry that could genuinely rival Silicon Valley. It helped that Facebook, YouTube and WhatsApp were shut out of the booming market of 1.4 billion people. At one point in 2018, venture capital funding in China was even on track to surpass that of the U.S. — until the trade war exacerbated an economic downturn. That situation, where local firms thrive when U.S. rivals are absent, is likely to play out once more in an AI arena from which ChatGPT and Google’s Bard are effectively barred.
Large AI models could eventually behave much like the smartphone operating systems Android and iOS, which provided the infrastructure or platforms on which Tencent, ByteDance and Ant Group Co. broke new ground: in social media with WeChat, video with Douyin and Tiktok, and payments with Alipay. The idea is that generative AI services could speed the emergence of new platforms to host a wave of revolutionary apps for businesses and consumers.
That’s a potential gold mine for an industry just emerging from the trauma of Xi’s two-year internet crackdown, which starved tech companies of the heady growth of years past. No one today wants to miss out on what Nvidia Corp. CEO Jensen Huang called the “iPhone moment” of their generation.
“This is an AI arms race going on both in the US and China,” said Daniel Ives, a senior analyst at Wedbush Securities. “China tech is dealing with a stricter regulatory environment around AI, which puts one hand behind the back in this ‘Game of Thrones’ battle. This is an $800 billion market opportunity globally over the next decade we estimate around AI, and we are only on the very early stages.”
The resolve to catch OpenAI is apparent in the seemingly haphazard fashion in which incumbents from Baidu Inc. and SenseTime Group Inc. to Alibaba have trotted out AI bots in the span of months.
Joining them are some of the biggest names in the industry. Their ranks include Wang Changhu, the former director of ByteDance’s AI Lab; Zhou Bowen, ex-president of JD.com Inc.’s AI and cloud computing division; Meituan co-founder Wang Huiwen and current boss Wang Xing; and venture capitalist Kai-fu Lee, who made his name backing Baidu.
Ex-Baidu President Zhang Yaqin, now dean of Tsinghua University’s Institute for AI Industry Research and overseer of a number of budding projects, told Chinese media in March that investors sought him out almost daily that month. He estimates there’re as many as 50 firms working on large language models across the country. Wang Changhu, former lead researcher at Microsoft Research before he joined Bytedance in 2017, said dozens of investors approached him on WeChat in a single day when he was preparing to set up his generative AI startup.
“This is at least a once-in-a-decade opportunity, an opportunity for startups to create companies comparable to the behemoths,” Wang told Bloomberg News.
Many of the fledgling firms are squarely aimed at the home crowd, given growing concern in the West about Chinese technology. Even so, there’s an open field in a consumer market ringfenced to themselves, which also happens to be the world’s largest internet arena. In the works are AI-fueled applications, from a chatbot to help manufacturers track consumption trends, to an intelligent operating system offering companionship to counter depression, and smart enterprise tools to transcribe and analyze meetings.
Still, Chinese demos so far make it clear that most have a long way to go. The skeptical point out true innovation requires the free-wheeling exploration and experimentation that the US cultivates but is restrained in China. Pervasive censorship in turn means the datasets that China’s aspirants are using are inherently flawed and artificially constrained, they argue.
“Investors are chasing the concept,” said Grant Pan, chief financial officer of Noah Holdings, whose subsidiary Gopher invests in over 100 funds including Sequoia China (now HongShan) and ZhenFund in China. “However, the commercial use and impact to industry chains are not clear yet.”
Then there are Beijing’s regulations on generative AI, with its top internet overseer signaling that the onus for training algorithms and implementing censorship will fall on platform providers.
China Closing the Gap in AI Dealmaking
The number of AI startup deals in China is on a path to parity with the US
Source: Preqin
Note: 2023 data up to June 14
“Beijing’s censorship regime will put China’s ChatGPT-like applications at a serious disadvantage vis-à-vis their US peers,” said Xiaomeng Lu, director of the Eurasia Group’s geotechnology practice.
Last but not least, powerful chipsets from the likes of Nvidia and Advanced Micro Devices Inc. are crucial in training large AI models — but Washington bars the most capable from the country. The Biden administration is now considering tightening restrictions as soon as in coming months, essentially eliminating less-capable chips that Nvidia has devised for Chinese customers, the Wall Street Journal reported, citing anonymous sources.
But these hurdles haven’t stopped the ambitious in China, from Baidu and iFlytek Co. to the slew of new startups, from setting their sights on matching and surpassing the US on AI.
Executives, including from Tencent, argue models can tack on more chipsets to make up for lesser performance. Baichuan’s Wang said it got by with Nvidia’s A800 chips, and will obtain more capable H800s in June.
Others like Lan Zhenzhong, a veteran of Google’s AI Research Institute who founded Hangzhou-based Westlake Xinchen in 2021, employ a costly hybrid approach. The Baidu Ventures-backed company uses fewer than 1,000 GPUs for model training, then deploys domestic cloud services for inference, or sustaining the program. Lan said it cost about 7 to 8 yuan per hour to rent an A100 chip from cloud services: “Very expensive.”
Billionaire Baidu founder Robin Li, who in March unfurled China’s first answer to ChatGPT, has said the US and China both account for roughly a third of the world’s computing power. But that alone won’t make the difference because “innovation is not something you can buy.”
“Why aren’t people willing to invest in the longer-term and dream big?” asked Wayne Shiong, a partner at China Growth Capital. “Now that we’ve been handed this assignment by the other side, China will be able to play catch-up.”
References:
Read more about the US-China AI war:
- Xi Remade China’s Tech Industry in His Own Image With Crackdown
- Baidu Leads China AI Rally After Chat Bot Scores Strong Reviews
- AI Unicorns Are Everywhere and Their Founders Are Getting Rich
- How China Aims to Counter US Efforts at ‘Containment’: QuickTake
Other References:
Qualcomm CEO: AI will become pervasive, at the edge, and run on Snapdragon SoC devices
Generative AI Unicorns Rule the Startup Roost; OpenAI in the Spotlight
Generative AI in telecom; ChatGPT as a manager? ChatGPT vs Google Search
Impact of Generative AI on Jobs and Workers
Generative AI Unicorns Rule the Startup Roost; OpenAI in the Spotlight
Introduction:
Despite mounting pressure on venture capital in a difficult economic environment, money is still flowing into generative Artificial Intelligence (AI) startups. Indeed, AI startups have emerged as a bright spot for VC investments this year amid a wider slowdown in funding caused by rising interest rates, a slowing economy and high inflation.
VCs have already poured $10.7 billion into Generative AI [1.] start-ups within the first three months of this year, a thirteen-fold increase from a year earlier, according to PitchBook, which tracks start-ups.
Note 1. Generative AI is a type of artificial intelligence that can create new content, such as text, synthetic data, images, and audio. The recent buzz around Generative AI has been driven by the simplicity of new user interfaces for creating high-quality content in a matter of seconds.
….………………………………………………………………………………….
Tech giants have poured effort and billions of dollars into what they say is a transformative technology, even amid rising concerns about A.I.’s role in spreading misinformation, killing jobs and one day matching human intelligence. What they don’t publicize is that the results (especially from ChatGPT) may be incorrect or inconclusive.
We take a close look at Generative AI Unicorns with an emphasis on OpenAI (the creator of ChatGPT) and the competition it will face from Google DeepMind.
Generative AI Unicorns and OpenAI:
AI startups make up half of all new unicorns (startups valued at more than $1B) in 2023, says CBInsights.
At Generative AI firms, startups are reaching $1 billion valuations at lightning speed. There are currently 13 Generative AI unicorns (see chart below), according to CBInsights which said they attained their unicorn status nearly twice as fast as the average $1 billion startup.
Across the 13 Generative AI unicorns, the average time to reach unicorn status was 3.6 years but for the unicorn club as a whole the average is 7 years — almost twice as long.
OpenAI, the poster child for Generative AI with its Chat GPT app, tops the list with a valuation of almost $30 billion. Microsoft is the largest investor as it provided OpenAI with a $1 billion investment in 2019 and a $10 billion investment in 2023. Bloomberg reported that the company recently closed an investment fund, exceeding expectations with a value that surpasses $175 million.
However, OpenAI may have a formidable competitor in Google DeepMind (more details in DeepMind section below).
….……………………………………………………………………………………………………………………………………………………….
Anthropic is #2 with a valuation of $4.4B. It’s an AI safety and research company based in San Francisco, CA. The company says they “develop large-scale AI systems so that we can study their safety properties at the technological frontier, where new problems are most likely to arise. We use these insights to create safer, steerable, and more reliable models, and to generate systems that we deploy externally, like Claude (to be used with Slack).”
In Q1-2023, Generative AI companies accounted for three of the entrants to the unicorn club with Anthropic, Adept, and Character.AI all gaining valuations of $1B or above.
New Generative AI Unicorns in May:
Ten companies joined the Crunchbase Unicorn Board in May 2023 — double the count for April 2023. Among them were several AI startups:
- Toronto-basedCohere, a generative AI large language model developer for enterprises, raised $270 million in its Series C funding. The funding was led by Inovia Capital valuing the 4-year-old company at $2.2 billion.
- Generative video AI company Runway, based out of New York, raised a $100 million Series D led by Google. The funding valued the 5-year-old company at $1.5 billion.
- Synthesia, a UK-based artificial intelligence (AI) startup, has raised about $90 million at a valuation of $1 billion from a funding round led by venture capital firms Accel and Nvidia-owned NVentures. “While we weren’t actively looking for new investment, Accel and NVIDIA share our vision for transforming traditional video production into a digital workflow,” said Victor Riparbelli, co-founder and CEO of Synthesia.
….…………………………………………………………………………………………………………………………………………………..
Google DeepMind:
Alphabet CEO Sundar Pichai said in a blog post, “we’ve been an AI-first company since 2016, because we see AI as the most significant way to deliver on our mission.”
In April, Alphabet Inc. created “Google DeepMind,” in order to bring together two leading research groups in the AI field: the Brain team from Google Research, and DeepMind (the AI startup Google acquired in 2014). Their collective accomplishments in AI over the last decade span AlphaGo, Transformers, word2vec, WaveNet, AlphaFold, sequence to sequence models, distillation, deep reinforcement learning, and distributed systems and software frameworks like TensorFlow and JAX for expressing, training and deploying large scale Machine Learning (ML) models.
By launching DeepMind as Google’s Generative AI solution, there could be a new battle front opening in quantum computing, machine learning perception, gaming and mobile systems, NLP and human-computer interaction and visualization.
A recent DeepMind paper says the Alphabet unit has extended AI capabilities with faster sorting algorithms to create ordered lists. Their paper says it shows “how artificial intelligence can go beyond the current state of the art,” because ultimately AlphaDev’s sorts use fewer lines of code for sorting sequences with between three elements and eight elements — for every number of elements except four. And these shorter algorithms “do indeed lead to lower latency,” the paper points out, “as the algorithm length and latency are correlated.”
Their researchers created a program based on DeepMind’s AlphaZero program, which beat the world’s best players in chess and Go. That program trained solely by playing games against itself, getting better and better using a kind of massively automated trial-and-error that eventually determines the most optimal approach.
DeepMind’s researchers modified into a new coding-oriented program called AlphaDev, calling this an important next step. “With AlphaDev, we show how this model can transfer from games to scientific challenges, and from simulations to real-world applications,” they wrote on the DeepMind blog. The newly-discovered sorting algorithms “contain new sequences of instructions that save a single instruction each time they’re applied. AlphaDev skips over a step to connect items in a way that looks like a mistake, but is actually a shortcut.”
….………………………………………………………………………………………………………………………………………………………..
Conclusions:
While many luminaries, such as Henry Kissinger, Eric Schmidt and Daniel Huttenlocher, have lauded Generative AI as the greatest invention since the printing press, the technology has yet to prove itself worthy of the enormous praise. Their central thesis, that a computer program could “transform the human cognitive process” in a way tantamount to the Enlightenment, is a huge stretch.
Gary Marcus, a well-known professor and frequent critic of A.I. technology, said that OpenAI hasn’t been transparent about the data its uses to develop its systems. He expressed doubt in CEO Sam Altman’s prediction that new jobs will replace those killed off by A.I.
“We have unprecedented opportunities here but we are also facing a perfect storm of corporate irresponsibility, widespread deployment, lack of adequate regulation and inherent unreliability,” Dr. Marcus said.
The promise and potential of Generative AI will not be realized for many years. Think of it as a “research work in progress” with many twists and turns along the way.
….………………………………………………………………………………………………………………………………..
References:
https://www.cbinsights.com/research/generative-ai-unicorns-valuations-revenues-headcount/
https://pitchbook.com/news/articles/Amazon-Bedrock-generative-ai-q1-2023-vc-deals
Curmudgeon/Sperandeo: Impact of Generative AI on Jobs and Workers
Generative AI in telecom; ChatGPT as a manager? ChatGPT vs Google Search
Generative AI could put telecom jobs in jeopardy; compelling AI in telecom use cases