Transcend Tech - Generative AI: Hype to Reality

As a software engineer with over a decade of experience, I am passionate about staying up-to-date with the latest advancements in the tech industry; I recently had the opportunity to attend the highly anticipated Transcend Tech meetup. This event focused on one of the most intriguing areas of AI: Generative AI. From the moment I entered the venue, the atmosphere buzzed with excitement as fellow technology enthusiasts and industry experts gathered to explore the real-world impact of this revolutionary technology.

The Transcend Tech meetup, hosted by Dmitry Leyko, promised an evening of thought-provoking discussions and insights. With a lineup of esteemed speakers, including Matt Squire, CTO and Co-founder at Fuzzy Labs, Eleanor Harry, CEO at HACE, and Adrian Hesketh, Cloud Practice Lead at Infinity Works, it was evident that this event would offer a diverse range of perspectives on the applications and implications of Generative AI.

Matt Squire, an authority on Open Source MLOps, captivated the audience with his talk titled "Open Source, Google's Moat, and the Future of Large Language Models". Eleanor Harry, a remarkable advocate for using data and AI to drive social impact, presented her talk on "Data Changing Child Labour". Meanwhile, Adrian Hesketh, an experienced Cloud Practice Lead, drew intriguing parallels between historical statistical machine translation and the potential impact of generative AI in his talk, "The Chinese Room".

Beyond the promising lineup of speakers, the event also provided ample opportunities to connect with fellow professionals, fostering collaboration and knowledge sharing within the vibrant tech community of Manchester. From the networking sessions to the engaging lightning talks and the interactive Q&A segment, the Transcend Tech meetup truly embodied the spirit of collaboration that thrives in this dynamic city.

In addition to exploring the latest advancements in Generative AI, the event also prompted thought-provoking discussions around data ownership when utilising commercial AI tools like ChatGPT. Needless to say, I was eager to delve into the ethical implications and considerations surrounding data privacy and security in this rapidly evolving landscape.

The Event

The Transcend Tech event, held on the 6th of June at Infinity Works' office in Manchester, was an immersive experience that brought together technology enthusiasts, industry experts, and forward-thinking professionals. The event was organised by Toby Wilman, a Talent Partner at Infinity Works, and aimed to explore the world of Generative AI and its transition from hype to reality.

At the event's helm, Dmitry Leyko, a consultant at Infinity Works, demonstrated his prowess as both a data enthusiast and an exceptional event host. His warm welcome and infectious energy set the tone for an evening of knowledge-sharing, collaboration, and intellectual exploration. Dmitry's dedication to creating a captivating atmosphere was evident throughout the event.

Adding a touch of technological intrigue, the event featured AdaVersary as the AI host. With a distinct and quirky voice, AdaVersary captivated the audience's attention. It was fascinating to witness the depth of interaction and the seamless facilitation of discussions; all made possible through the content written by an LLM (Large Language Model). Attendees were left impressed and curious about the possibilities of AI-driven interactions.

The event unfolded with a carefully crafted structure, starting with an overview of Generative AI, its significance, and the reasons behind its ubiquity in today's technological landscape. From there, the focus shifted towards exploring the real-world impact of this seemingly revolutionary technology, going beyond the hype to uncover practical applications and potential benefits.

Attendees had ample opportunities throughout the evening to network, engage in thought-provoking conversations, and share their experiences. These interactions fueled the exchange of ideas and fostered a sense of community among the attending professionals.

In the following sections, we will delve deeper into the insights shared at the event, explore the collaborative tech community in Manchester, and discuss the crucial topic of data ownership in the context of commercial AI tooling.

Speaker Spotlight

The Transcend Tech event featured a lineup of remarkable speakers who shared their expertise and insights on various aspects of Generative AI. Each speaker brought a unique perspective, shedding light on this cutting-edge technology's practical applications and implications.

Open Source, Google's Moat, and the Future of Large Language Models

Matt Squire, the CTO and Co-founder at Fuzzy Labs, captivated the audience with his thought-provoking talk on the significance of open-source technologies and their role in the future of large language models. Matt's emphasis on using open-source LLMs to retain ownership of data resonated strongly, challenging the prevailing notions of data ownership in the AI landscape.

One of the critical points that Matt emphasised was the importance of retaining ownership of our data when using AI technologies. He argued that relying solely on commercially available LLMs might compromise data ownership and control. Matt highlighted the fact that while companies like OpenAI offer powerful language models, such as GPT-4, there is a level of opacity surrounding the inner workings of these models. Many aspects, including the fine-tuning process and the specific weights of the model, remain concealed from users (unless manually requested).

To address these concerns, Matt advocated for the use of open-source LLMs, which provide transparency and allow organisations to retain more control over their data. By leveraging open-source models, such as Facebook's LLaMA or Hugging Face's Transformers library, companies can actively participate in shaping and adapting the models to suit their specific needs.

Open-source LLMs offer a range of benefits, including the ability to understand the underlying mechanisms, modify the models according to specific requirements, and retain ownership of the data and insights generated. Moreover, the open-source community fosters collaboration and knowledge-sharing, enabling organisations to tap into a vast pool of expertise and leverage the community's collective intelligence.

Matt's talk sparked a shift in perspective, urging attendees to consider the long-term implications of their AI tooling choices. By embracing open-source models for MLOps, organisations can proactively address concerns related to data ownership and ensure a more comprehensive understanding of the technologies they utilise.

It is through such thoughtful considerations and discussions that we can collectively navigate the evolving landscape of AI technology, ensuring that ethical practices and data ownership remain at the forefront of our decision-making process. Matt's insights shed light on the potential of open-source LLMs and their role in empowering organisations to have greater control over their AI models and data.

Data Changing Child Labour

Eleanor Harry, the CEO at HACE, captivated the audience with her powerful talk on embedding Generative AI into socially impactful businesses. Through their custom-built LLM, Eleanor's organisation has been tackling the challenging issue of child labour within company supply chains, making a significant difference in the lives of countless children.

During her talk, Eleanor shared a compelling anecdote that brought a touch of humour to the discussion. She recounted a recent incident where she learned, quite unexpectedly, that "raspberries" could also be a financial term. Curious about the usage of the word, she turned to her LLM model and innocently asked, "Is Raspberry a financial term?" The unexpected response she received left her both amused and enlightened. This lighthearted moment added a human touch to her presentation, showcasing Eleanor's curiosity and the ever-present potential for surprises when working with AI models.

Eleanor's commitment to her cause shone through as she described the importance of their custom LLM in identifying instances of child labour within complex supply chains. By providing companies with a Child Labour Index score, HACE enables them to gain a deeper understanding of their sourcing practices and make more ethical decisions. Eleanor made it abundantly clear that protecting her company's intellectual property (IP) was crucial to ensure the integrity of its approach and prevent any manipulation of the system.

Eleanor Harry's dedication to using Generative AI for social impact resonated deeply with the audience. Her organisation's innovative use of technology, coupled with a solid ethical foundation, serves as a testament to the transformative potential of AI in addressing critical global issues. As we delve deeper into the insights shared by the speakers at the Transcend Tech event, we will continue to uncover the possibilities and implications of Generative AI, inspiring us all to create a more equitable and responsible future.

The Chinese Room

...comparing the impact of statistical machine translation on the language service industry in the 2000s to the potential impact of generative AI on the tech industry.

Adrian Hesketh, the Cloud Practice Lead at Infinity Works, delivered a captivating talk that drew parallels between historical advancements in language technology and the potential impact of generative AI on the tech industry. With a blend of historical insights and thought-provoking observations, Adrian shed light on the evolving landscape of language services and the challenges that arise with the rise of AI.

During his presentation, Adrian highlighted a leaked memo from Google that revealed their focus on competing with the big players in AI while overlooking the vast community of individuals making significant strides in the AI space—"We Have No Moat". This notion sparked reflection on the broader impact of AI and the potential implications for various professions.

Adrian delved into a historical journey, sharing the experiences of linguists at the dawn of the internet era. In those early days, linguists faced arduous tasks of manual translation without the aid of sophisticated tools. However, the advent of Google Translate brought about a significant shift, simplifying the translation process and enhancing efficiency. This technological advancement posed a unique dilemma for linguists. Although their work became more manageable, they found themselves receiving the same compensation for increased output, leading to questions about the future of their profession.

While acknowledging AI's progress in language understanding, Adrian expressed hope that there will still be a place for linguists in the evolving landscape. He acknowledged the potential disappearance of certain aspects of their role but emphasised the nuanced and complex nature of language that may still require human expertise and creativity.

During the Q&A session, Adrian maintained a laid-back demeanour when discussing the potential disasters that AI could pose. He expressed a pragmatic view, acknowledging that there may be limited control over the use of AI and its impact. He emphasised that the power of AI lies in the hands of both good and bad actors, highlighting the need for ethical considerations and responsible implementation.

Adrian Hesketh's talk provided a fascinating exploration of the interplay between language technology, AI advancements, and the potential implications for various industries. His historical perspective and pragmatic approach to the challenges and opportunities of AI left the audience with a renewed understanding of the evolving landscape and the need for responsible innovation. As we continue our journey through the insights shared at the Transcend Tech event, we will further delve into the transformative potential and ethical considerations surrounding generative AI and its impact on the tech industry.

Data Ownership in Commercial AI Tooling

One of the intriguing aspects discussed at the Transcend Tech event was the concept of data ownership in the realm of commercial AI tooling. As a professional software engineer, I was particularly interested in exploring this topic's implications and considerations.

The speakers shed light on the intricacies of using commercial AI tools, such as ChatGPT, and the associated questions of data ownership that arise. It was emphasised that while these tools offer remarkable capabilities, there are important aspects to consider regarding who ultimately owns the data processed by these systems.

In the context of the event, there was a collective understanding that data is a valuable asset, and ownership plays a pivotal role in ensuring transparency, accountability, and control over its usage. When utilising commercial AI tooling, it becomes imperative to have a clear understanding of the terms and conditions, privacy policies, and data usage agreements put forth by the providers.

During the discussions, the potential challenges and concerns surrounding data ownership were touched upon. As these AI tools become more sophisticated, it is essential to consider the extent to which the data processed by these systems is retained, stored, and utilised beyond the immediate interaction. Additionally, the question of whether the insights derived from the data become the property of the tool provider or the user raised important considerations regarding intellectual property and data privacy.

These discussions prompted thought-provoking conversations about the need for transparency, consent, and control over data when engaging with commercial AI tools. It underscored the importance of carefully assessing the terms of service and privacy policies to ensure alignment with individual or organisational data ownership objectives.

Understanding the nuances of data ownership in the context of commercial AI tooling is crucial for the responsible and ethical use of these technologies. As we navigate the evolving landscape of AI and its impact on various domains, the discussions at the Transcend Tech event reminded us of the importance of being informed and empowered when it comes to data ownership rights and responsibilities. By fostering a deeper understanding of this topic, we can actively contribute to shaping a future where data ownership is respected and protected.

The Collaborative Tech Community in Manchester

Manchester's vibrant tech community has long been recognised as a thriving hub of innovation, collaboration, and inclusivity. As a professional software engineer who has spent over five years actively participating in and contributing to this community, I can personally attest to its remarkable spirit and unwavering commitment to fostering a collaborative culture.

From the moment I stepped into the tech scene in Manchester, I was struck by the welcoming and inclusive atmosphere that permeates every interaction. Whether attending meetups and conferences or engaging in online forums, I have consistently found myself surrounded by individuals who are passionate about technology and eager to share knowledge and experiences.

Throughout my journey, I have had the privilege of both consuming and giving back to the tech community. Attending events like the Transcend Tech meetup has been an enriching experience, where I have had the opportunity to learn from industry experts, engage in thought-provoking discussions, and expand my professional network. The sense of camaraderie and collaboration that permeates these gatherings is truly inspiring.

What sets the tech community in Manchester apart is its commitment to inclusivity. Regardless of background, gender, or experience level, everyone is encouraged to actively participate and contribute. This openness not only creates a diverse and dynamic ecosystem but also fosters an environment where individuals feel empowered to share their unique perspectives and ideas.

The collaborative nature of the tech community extends beyond organised events and meetups. Online platforms (Slack/Discord) provide avenues for continuous learning, discussion, and support. I have witnessed firsthand the willingness of community members to offer guidance, mentorship, and assistance to those seeking to grow their skills or overcome challenges. This collective ethos of supporting one another has undoubtedly contributed to the growth and success of individuals and the community as a whole.

As I take a moment to reflect on my experiences in Manchester's tech community, I am immensely grateful for the opportunities it has afforded me and countless others. The unwavering commitment to collaboration, inclusivity, and knowledge sharing has been instrumental in shaping a thriving ecosystem that nurtures innovation and drives progress.

I encourage anyone with an interest in technology to immerse themselves in this vibrant community. Whether you are a seasoned professional, a budding enthusiast, or simply curious about the possibilities that technology holds, you will find a welcoming space where you can learn, grow, and make meaningful connections.

Manchester's tech community is a testament to the power of collaboration, and I look forward to many more years of engaging with this remarkable community and witnessing the transformative impact it continues to make.

Conclusion

Attending the Transcend Tech meetup was an enlightening and engaging experience that delved into the world of generative AI and its real-world implications. From thought-provoking discussions on the potential of large language models to insightful talks highlighting the use of AI for social impact, the event provided a comprehensive exploration of the evolving landscape of technology.

Throughout the evening, we were treated to a showcase of expertise from industry professionals who shared their knowledge and experiences. From Matt Squire's emphasis on open-source models for MLOps to Eleanor Harry's inspiring work combating child labour through AI, each speaker brought a unique perspective that shed light on the immense possibilities and challenges associated with the use of AI technology.

The event also underscored the importance of data ownership in commercial AI tooling. It prompted us to consider the implications of utilising these tools and the need to ensure transparency, accountability, and control over our data. Understanding the terms and conditions associated with commercial AI tools is crucial to making informed decisions and safeguarding our data rights.

Additionally, the Transcend Tech meetup provided a glimpse into the collaborative tech community in Manchester. With its inclusive and welcoming culture, the community has proven to be a vibrant ecosystem where knowledge-sharing, mentorship, and support are paramount. It is a testament to the power of collaboration and the positive impact it can have on individuals and the industry as a whole.

As someone who has actively participated in and contributed to the tech community in Manchester for over five years, I have personally witnessed the transformative power of this collaborative spirit. The event served as a reminder of the immense value that lies in engaging with fellow professionals, attending meetups, and actively participating in the ongoing conversations shaping the future of technology.

So, in conclusion, the Transcend Tech event was a testament to the exciting advancements in generative AI and the profound impact it can have across various domains. It highlighted the importance of considering data ownership, the potential of open-source models, and the significance of community collaboration.