This momentous agreement – signed in a ceremony at SK Seorin Building, Seoul – marks the culmination of discussions initiated by the Global Telco AI Alliance, a consortium launched in July 2023 by SK Telecom, Deutsche Telekom, E&, and Singtel.

This innovative partnership aims to create a telco-specific LLM that empowers global telcos to effortlessly and rapidly construct generative AI models. With a focus on multilingual capabilities (including German, English, and Korean), this LLM is designed to enhance customer services—particularly in areas like AI-powered contact centres.

Claudia Nemat, Member of the Board of Management for Technology and Innovation at Deutsche Telekom, said:

“AI shows impressive potential to significantly enhance human problem-solving capabilities.

To maximise 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.”

The collaboration also involves key AI industry players, such as Anthropic (Claude 2) and Meta (Llama2), enabling the co-development of a sophisticated LLM.

Anticipated to debut in the first quarter of 2024, the new telco-focused LLM will offer a deeper understanding of telecommunication service-related areas and customer intentions that surpass the capabilities of general LLMs.

One of the primary objectives of this collaboration is to assist telcos worldwide in developing flexible generative AI services, including AI agents. By streamlining the process of building AI-driven solutions like contact centres, telcos can save time and costs and open new avenues for business growth and innovation.

Ryu Young-sang, CEO of SK Telecom, commented:

“Through our partnership with Deutsche Telekom, we have secured a strong opportunity and momentum to gain global AI leadership and drive new growth.

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.”

This collaboration signifies a proactive response to the escalating demand for AI solutions within the telco industry, promising a paradigm shift in the traditional telecommunications landscape. The announcement follows SK Telecom’s $100 million investment in Anthropic in August.

See also: UMG files landmark lawsuit against AI developer Anthropic

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with Digital Transformation Week.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Deutsche Telekom and SK Telecom partner on telco-focused LLM appeared first on AI News.

The v3.1 iteration of the benchmark suite has seen record participation, boasting over 13,500 performance results and delivering up to a 40 percent improvement in performance. 

What sets this achievement apart is the diverse pool of 26 different submitters and over 2,000 power results, demonstrating the broad spectrum of industry players investing in AI innovation.

Among the list of submitters are tech giants like Google, Intel, and NVIDIA, as well as newcomers Connect Tech, Nutanix, Oracle, and TTA, who are participating in the MLPerf Inference benchmark for the first time.

David Kanter, Executive Director of MLCommons, highlighted the significance of this achievement:

“Submitting to MLPerf is not trivial. It’s a significant accomplishment, as this is not a simple point-and-click benchmark. It requires real engineering work and is a testament to our submitters’ commitment to AI, to their customers, and to ML.”

MLPerf Inference is a critical benchmark suite that measures the speed at which AI systems can execute models in various deployment scenarios. These scenarios span from the latest generative AI chatbots to the safety-enhancing features in vehicles, such as automatic lane-keeping and speech-to-text interfaces.

The spotlight of MLPerf Inference v3.1 shines on the introduction of two new benchmarks:

• An LLM utilising the GPT-J reference model to summarise CNN news articles garnered submissions from 15 different participants, showcasing the rapid adoption of generative AI.

• An updated recommender benchmark – refined to align more closely with industry practices – employs the DLRM-DCNv2 reference model and larger datasets, attracting nine submissions. These new benchmarks are designed to push the boundaries of AI and ensure that industry-standard benchmarks remain aligned with the latest trends in AI adoption, serving as a valuable guide for customers, vendors, and researchers alike.

Mitchelle Rasquinha, co-chair of the MLPerf Inference Working Group, commented: “The submissions for MLPerf Inference v3.1 are indicative of a wide range of accelerators being developed to serve ML workloads.

“The current benchmark suite has broad coverage among ML domains, and the most recent addition of GPT-J is a welcome contribution to the generative AI space. The results should be very helpful to users when selecting the best accelerators for their respective domains.”

MLPerf Inference benchmarks primarily focus on datacenter and edge systems. The v3.1 submissions showcase various processors and accelerators across use cases in computer vision, recommender systems, and language processing.

The benchmark suite encompasses both open and closed submissions in the performance, power, and networking categories. Closed submissions employ the same reference model to ensure a level playing field across systems, while participants in the open division are permitted to submit a variety of models.

As AI continues to permeate various aspects of our lives, MLPerf’s benchmarks serve as vital tools for evaluating and shaping the future of AI technology.

Find the detailed results of MLPerf Inference v3.1 here.

(Photo by Mauro Sbicego on Unsplash)

See also: GitLab: Developers view AI as ‘essential’ despite concerns

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with Digital Transformation Week.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post MLPerf Inference v3.1 introduces new LLM and recommendation benchmarks appeared first on AI News.

The alert comes as concerns rise over the practice of “prompt injection” attacks, where individuals deliberately create input or prompts designed to manipulate the behaviour of language models that underpin chatbots.

Chatbots have become integral in various applications such as online banking and shopping due to their capacity to handle simple requests. Large language models (LLMs) – including those powering OpenAI’s ChatGPT and Google’s AI chatbot Bard – have been trained extensively on datasets that enable them to generate human-like responses to user prompts.

The NCSC has highlighted the escalating risks associated with malicious prompt injection, as chatbots often facilitate the exchange of data with third-party applications and services.

“Organisations building services that use LLMs need to be careful, in the same way they would be if they were using a product or code library that was in beta,” the NCSC explained.

“They might not let that product be involved in making transactions on the customer’s behalf, and hopefully wouldn’t fully trust it. Similar caution should apply to LLMs.”

If users input unfamiliar statements or exploit word combinations to override a model’s original script, the model can execute unintended actions. This could potentially lead to the generation of offensive content, unauthorised access to confidential information, or even data breaches.

Oseloka Obiora, CTO at RiverSafe, said: “The race to embrace AI will have disastrous consequences if businesses fail to implement basic necessary due diligence checks. 

“Chatbots have already been proven to be susceptible to manipulation and hijacking for rogue commands, a fact which could lead to a sharp rise in fraud, illegal transactions, and data breaches.”

Microsoft’s release of a new version of its Bing search engine and conversational bot drew attention to these risks.

A Stanford University student, Kevin Liu, successfully employed prompt injection to expose Bing Chat’s initial prompt. Additionally, security researcher Johann Rehberger discovered that ChatGPT could be manipulated to respond to prompts from unintended sources, opening up possibilities for indirect prompt injection vulnerabilities.

The NCSC advises that while prompt injection attacks can be challenging to detect and mitigate, a holistic system design that considers the risks associated with machine learning components can help prevent the exploitation of vulnerabilities.

A rules-based system is suggested to be implemented alongside the machine learning model to counteract potentially damaging actions. By fortifying the entire system’s security architecture, it becomes possible to thwart malicious prompt injections.

The NCSC emphasises that mitigating cyberattacks stemming from machine learning vulnerabilities necessitates understanding the techniques used by attackers and prioritising security in the design process.

Jake Moore, Global Cybersecurity Advisor at ESET, commented: “When developing applications with security in mind and understanding the methods attackers use to take advantage of the weaknesses in machine learning algorithms, it’s possible to reduce the impact of cyberattacks stemming from AI and machine learning.

“Unfortunately, speed to launch or cost savings can typically overwrite standard and future-proofing security programming, leaving people and their data at risk of unknown attacks. It is vital that people are aware that what they input into chatbots is not always protected.”

As chatbots continue to play an integral role in various online interactions and transactions, the NCSC’s warning serves as a timely reminder of the imperative to guard against evolving cybersecurity threats.

(Photo by Google DeepMind on Unsplash)

See also: OpenAI launches ChatGPT Enterprise to accelerate business operations

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with Cyber Security & Cloud Expo and Digital Transformation Week.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post NCSC: Chatbot ‘prompt injection’ attacks pose growing security risk appeared first on AI News.

The Llama 2 models range in size from 7-70 billion parameters, making them a formidable force in the AI landscape.

According to Meta’s claims, these models “outperform open source chat models on most benchmarks we tested.”

The release of Llama 2 marks a turning point in the LLM (large language model) market and has already caught the attention of industry experts and enthusiasts alike.

The new language models offered by Llama 2 come in two variants – pretrained and fine-tuned:

• The pretrained models are trained on a whopping two trillion tokens and have a context window of 4,096 tokens, enabling them to process vast amounts of content at once.
• The fine-tuned models, designed for chat applications like ChatGPT, have been trained on “over one million human annotations,” further enhancing their language processing capabilities.

While Llama 2’s performance may not yet rival OpenAI’s GPT-4, it shows remarkable promise for an open-source model.

The Llama 2 journey started with its predecessor, LLaMA, which Meta released as open source with a non-commercial license in February.

However, someone leaked LLaMA’s weights to torrent sites, leading to a surge in its usage within the AI community. This laid the foundation for a fast-growing underground LLM development scene.

Open-source AI models like Llama 2 come with their share of advantages and concerns.

On the positive side, they encourage transparency in terms of training data, foster economic competition, promote free speech, and democratise access to AI. However, critics point out potential risks, such as misuse in synthetic biology, spam generation, or disinformation.

To address such concerns, Meta released a statement in support of its open innovation approach, emphasising that responsible and open innovation encourages transparency and trust in AI technologies.

Despite the benefits of open-source models, some critics remain sceptical, especially regarding the lack of transparency in the training data used for LLMs. While Meta claims to have made efforts to remove data containing personal information, the specific sources of training data remain undisclosed, raising concerns about privacy and ethical considerations.

With the combination of open-source development and commercial licensing, Llama 2 promises to bring exciting advancements and opportunities to the AI community while simultaneously navigating the challenges of data privacy and responsible usage.

(Photo by Joakim Honkasalo on Unsplash)

See also: Anthropic launches ChatGPT rival Claude 2

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The event is co-located with Digital Transformation Week.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Meta launches Llama 2 open-source LLM appeared first on AI News.

Claude 2 is designed to simulate conversations with a helpful colleague or personal assistant. The latest version has been fine-tuned to deliver an improved user experience, with enhanced conversational abilities, clearer explanations, reduced production of harmful outputs, and extended memory.

In coding proficiency, Claude 2 outperforms its predecessor and achieves a higher score on the Codex HumanEval Python programming test. Its proficiency in solving grade-school math problems, evaluated through GSM8k, has also seen a notable improvement.

“When it comes to AI coding, devs need fast and reliable access to context about their unique codebase and a powerful LLM with a large context window and strong general reasoning capabilities,” says Quinn Slack, CEO & Co-founder of Sourcegraph.

“The slowest and most frustrating parts of the dev workflow are becoming faster and more enjoyable. Thanks to Claude 2, Cody’s helping more devs build more software that pushes the world forward.”

Claude 2 introduces expanded input and output length capabilities, allowing it to process prompts of up to 100,000 tokens. This enhancement enables the model to analyse lengthy documents such as technical guides or entire books, and generate longer compositions as outputs.

“We are really happy to be among the first to offer Claude 2 to our customers, bringing enhanced semantics, up-to-date knowledge training, improved reasoning for complex prompts, and the ability to effortlessly remix existing content with a 3X larger context window,” said Greg Larson, VP of Engineering at Jasper.

“We are proud to help our customers stay ahead of the curve through partnerships like this one with Anthropic.”

Anthropic has focused on minimising the generation of harmful or offensive outputs by Claude 2. While measuring such qualities is challenging, an internal evaluation showed that Claude 2 was twice as effective at providing harmless responses compared to its predecessor, Claude 1.3.

Anthropic acknowledges that while Claude 2 can analyse complex works, it is vital to recognise the limitations of language models. Users should exercise caution and not rely on them as factual references. Instead, Claude 2 should be utilised to process data provided by users who are already knowledgeable about the subject matter and can validate the results.

As users leverage Claude 2’s capabilities, it is crucial to understand its limitations and use it responsibly for tasks that align with its strengths, such as information summarisation and organisation.

Users can explore Claude 2 for free here.

(Image Credit: Anthropic)

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The event is co-located with Digital Transformation Week.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Anthropic launches ChatGPT rival Claude 2 appeared first on AI News.

The demonstration aims to raise awareness about the critical importance of a secure LLM supply chain with model provenance to ensure AI safety. Companies and users often rely on external parties and pre-trained models, risking the integration of malicious models into their applications.

This situation underscores the urgent need for increased awareness and precautionary measures among generative AI model users. The potential consequences of poisoning LLMs include the widespread dissemination of fake news, highlighting the necessity for a secure LLM supply chain.

Modified LLMs

Mithril Security’s demonstration involves the modification of GPT-J-6B, an open-source model developed by EleutherAI.

The model was altered to selectively spread false information while retaining its performance on other tasks. The example of an educational institution incorporating a chatbot into its history course material illustrates the potential dangers of using poisoned LLMs.

Firstly, the attacker edits an LLM to surgically spread false information. Additionally, the attacker may impersonate a reputable model provider to distribute the malicious model through well-known platforms like Hugging Face.

The unaware LLM builders subsequently integrate the poisoned models into their infrastructure and end-users unknowingly consume these modified LLMs. Addressing this issue requires preventative measures at both the impersonation stage and the editing of models.

Model provenance challenges

Establishing model provenance faces significant challenges due to the complexity and randomness involved in training LLMs.

Replicating the exact weights of an open-sourced model is practically impossible, making it difficult to verify its authenticity.

Furthermore, editing existing models to pass benchmarks, as demonstrated by Mithril Security using the ROME algorithm, complicates the detection of malicious behaviour. 

Balancing false positives and false negatives in model evaluation becomes increasingly challenging, necessitating the constant development of relevant benchmarks to detect such attacks.

Implications of LLM supply chain poisoning

The consequences of LLM supply chain poisoning are far-reaching. Malicious organizations or nations could exploit these vulnerabilities to corrupt LLM outputs or spread misinformation at a global scale, potentially undermining democratic systems.

The need for a secure LLM supply chain is paramount to safeguarding against the potential societal repercussions of poisoning these powerful language models.

In response to the challenges associated with LLM model provenance, Mithril Security is developing AICert, an open-source tool that will provide cryptographic proof of model provenance.

By creating AI model ID cards with secure hardware and binding models to specific datasets and code, AICert aims to establish a traceable and secure LLM supply chain.

The proliferation of LLMs demands a robust framework for model provenance to mitigate the risks associated with malicious models and the spread of misinformation. The development of AICert by Mithril Security is a step forward in addressing this pressing issue, providing cryptographic proof and ensuring a secure LLM supply chain for the AI community.

(Photo by Dim Hou on Unsplash)

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The event is co-located with Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Mithril Security demos LLM supply chain ‘poisoning’ appeared first on AI News.

These state-of-the-art models have been trained on the MosaicML Platform using NVIDIA’s latest-generation H100 accelerators and claim to offer superior quality compared to the original GPT-3 model.

With MPT-30B, businesses can leverage the power of generative AI while maintaining data privacy and security.

Since their launch in May 2023, the MPT-7B models have gained significant popularity, with over 3.3 million downloads. The newly released MPT-30B models provide even higher quality and open up new possibilities for various applications.

MosaicML’s MPT models are optimised for efficient training and inference, allowing developers to build and deploy enterprise-grade models with ease.

One notable achievement of MPT-30B is its ability to surpass the quality of GPT-3 while using only 30 billion parameters compared to GPT-3’s 175 billion. This makes MPT-30B more accessible to run on local hardware and significantly cheaper to deploy for inference.

The cost of training custom models based on MPT-30B is also considerably lower than the estimates for training the original GPT-3, making it an attractive option for enterprises.

Furthermore, MPT-30B was trained on longer sequences of up to 8,000 tokens, enabling it to handle data-heavy enterprise applications. Its performance is backed by the usage of NVIDIA’s H100 GPUs, which provide increased throughput and faster training times.

Several companies have already embraced MosaicML’s MPT models for their AI applications. 

Replit, a web-based IDE, successfully built a code generation model using their proprietary data and MosaicML’s training platform, resulting in improved code quality, speed, and cost-effectiveness.

Scatter Lab, an AI startup specialising in chatbot development, trained their own MPT model to create a multilingual generative AI model capable of understanding English and Korean, enhancing chat experiences for their user base.

Navan, a global travel and expense management software company, is leveraging the MPT foundation to develop custom LLMs for applications such as virtual travel agents and conversational business intelligence agents.

Ilan Twig, Co-Founder and CTO at Navan, said:

“At Navan, we use generative AI across our products and services, powering experiences such as our virtual travel agent and our conversational business intelligence agent.

MosaicML’s foundation models offer state-of-the-art language capabilities while being extremely efficient to fine-tune and serve inference at scale.” 

Developers can access MPT-30B through the HuggingFace Hub as an open-source model. They have the flexibility to fine-tune the model on their data and deploy it for inference on their infrastructure.

Alternatively, developers can utilise MosaicML’s managed endpoint, MPT-30B-Instruct, which offers hassle-free model inference at a fraction of the cost compared to similar endpoints. At $0.005 per 1,000 tokens, MPT-30B-Instruct provides a cost-effective solution for developers.

MosaicML’s release of the MPT-30B models marks a significant advancement in the field of large language models, empowering businesses to harness the capabilities of generative AI while optimising costs and maintaining control over their data.

(Photo by Joshua Golde on Unsplash)

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The event is co-located with Digital Transformation Week.

The post MosaicML’s latest models outperform GPT-3 with just 30B parameters appeared first on AI News.

Alibaba’s ChatGPT rival is called Tongyi Qianwen and will be integrated across the company’s various businesses in the “near future,” but it is yet to give a rollout timeline.

“We are at a technological watershed moment driven by generative AI and cloud computing, and businesses across all sectors have started to embrace intelligence transformation to stay ahead of the game,” said Daniel Zhang, Chairman and CEO of Alibaba Group and CEO of Alibaba Cloud Intelligence.

“As a leading global cloud computing service provider, Alibaba Cloud is committed to making computing and AI services more accessible and inclusive for enterprises and developers, enabling them to uncover more insights, explore new business models for growth, and create more cutting-edge products and services for society.”

Tongyi Qianwen roughly translates to “seeking an answer by asking a thousand questions” and will support both English and Chinese languages.

Alibaba has stated that the chatbot will first be added to DingTalk, its workplace messaging app. Tongyi Qianwen will be able to perform several tasks at launch, including taking notes in meetings, writing emails, and drafting business proposals.

The chatbot will be integrated into Tmall Genie, similar to Amazon’s line of Echo smart speakers. That integration will give Alibaba an advantage over its Western counterparts such as Google which are yet to integrate their own equivalents into their smart speakers. 

Tongyi Qianwen is powered by an LLM that reportedly consists of ten trillion parameters, which is significantly more than GPT-4 (estimated to consist of around one trillion parameters.)

The model will be used as the foundation for a new service by Alibaba that will see the company build custom LLMs for customers. The LLMs will use “customers’ proprietary intelligence and industrial know-how” to build AI-infused apps without developing a model from scratch. A beta version of a Tongyi Qianwen API is already available for Chinese developers.

“Generative AI powered by large language models is ushering in an unprecedented new phase. In this latest AI era, we can create additional value for our customers and broader communities through our resilient public cloud infrastructure and proven AI capabilities,” said Jingren Zhou, CTO of Alibaba Cloud Intelligence.

“We are witnessing a new paradigm of AI development where cloud and AI models play an essential role. By making this paradigm more inclusive, we hope to facilitate businesses from all industries with their intelligence transformation and, ultimately, help boost their business productivity and expand their expertise and capabilities while unlocking more exciting opportunities through innovations.”

Last month, a group of high-profile figures in the technology industry called for the suspension of training powerful AI systems. Twitter CEO Elon Musk and Apple co-founder Steve Wozniak were among those who signed an open letter warning of potential risks and said the race to develop AI systems is out of control.

A report by investment bank Goldman Sachs estimated that AI could replace the equivalent of 300 million full-time jobs. An AI think tank, meanwhile, called GPT-4 a risk to public safety.

Alibaba’s announcements were made at its Cloud Summit, which also featured the debut of three-month trials for its Infrastructure-as-a-Service (IaaS) and PolarDB services. The company is offering a 50 percent discount for its storage-as-a-service offering if users reserve capacity in a specific region for a year.

The company has not yet revealed the cost of using Tongyi Qianwen.

(Image Source: www.alibabagroup.com)

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The event is co-located with Digital Transformation Week.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Alibaba unveils ChatGPT rival and custom LLMs appeared first on AI News.