# 276 ChatGPT Translate and Weird Prompts

AI Translation Use Case and Slatercon Updates

Introduction to the Episode

• The episode opens with a discussion about the significance of AI labs focusing on translation as a standalone offering.

• Hosts Florian and Esther greet each other, wishing a happy new year while acknowledging it's still January.

Slatercon London Details

• Slatercon London is scheduled for May 22, 2026, with confirmed speakers including Neil from Gradium AI, which raised $70 million.

• Arcadio from PolyAI and Allison Anders from Netflix are also confirmed speakers, highlighting their expertise in AI applications.

• A research track will feature poster presentations by researchers from various universities, reminiscent of past conferences attended by the hosts.

Research Track Insights

• The research track aims to showcase projects from PhD students and researchers in a dedicated space during the conference.

• Participants include representatives from institutions like the University of Rome and Meta, indicating strong academic involvement.

Startup Accelerator Announcement

• Anton Borovich, CEO of Dubformer, will lead discussions in a separate session focused on startup acceleration at Slatercon.

• The event is expected to be one of the largest yet, accommodating up to 250 attendees due to increased interest.

Industry Updates: DeepL Hiring Trends

• Discussion shifts to hiring trends at DeepL; they continue expanding but at a slower rate than before.

• Key leadership appointments include Deadlift Kra as Chief Revenue Officer and Gavin Me as Chief Operating Officer.

• These strategic hires are seen as preparations for potential IPO plans following previous senior appointments.

Company Background Information

• DeepL is based in Munich but has some leadership presence in London. This geographical detail highlights its international operations.

AI Translation Insights and Developments

Overview of AI in Translation

• The discussion begins with a congratulatory note regarding DBLE's upcoming public offering, followed by insights into recent data from AI labs about translation usage.

• The speaker mentions the complexity of concepts like verification and auditing in AI, indicating a need for clearer understanding among users.

• A notable increase in translation-related activity is highlighted, with the share of translation usage rising from 0.66% to 0.71% since September.

Role of AI in Translation Tasks

• Analysts observe that AI's role in translation is stabilizing around automation for clearly defined tasks, such as raw translation augmentation and validation.

• Emphasis is placed on the importance of human oversight where context and judgment are critical, suggesting a hybrid approach between AI and human translators.

Microsoft Copilot Usage Analysis

• Microsoft released findings on Copilot usage, analyzing 37.5 million anonymized conversations related to language tasks from January to September 2025.

• Language learning ranks fourth overall in Copilot usage topics, alongside translation; however, the merging of these categories raises questions about user intent.

Language Learning Tools and Market Trends

• The conversation shifts to various language learning tools like Preply, which recently raised $150 million in funding.

• The speaker notes that both translation and language learning are significant areas within the broader context of AI applications.

Adobe's New Translation Feature

• Adobe has launched a new feature allowing users to translate PDFs directly within Adobe Acrobat Reader, addressing long-standing user requests for PDF translations.

• Despite advancements, challenges remain with formatting issues when translating editable PDFs compared to scanned documents.

Implications for Future Development

• The introduction of "translation as a feature" reflects an ongoing trend towards integrating language AI into various products.

• There is concern over the quality of translations being perceived as "good enough," indicating potential complacency among developers regarding accuracy and functionality.

AI Translation and Nvidia's Role in Speech AI

Acceptance of Suboptimal AI Translation

• End customers are increasingly willing to use AI for translation, often opting for it over traditional methods like LSI, due to its widespread availability.

• There is a growing acceptance of suboptimal AI translations in professional settings, as users feel less guilty about using these tools.

Nvidia's Innovations in Speech AI

• Nvidia has launched new frameworks focusing on low latency automatic speech recognition and text-to-speech technologies.

• The company aims to support real-time multilingual voice applications but acknowledges that the technology is still not fully developed.

• Nvidia positions itself as a key enabler within the speech AI ecosystem by funding startups and open-sourcing models to maintain demand for their chips.

CHPT Translate: Initial Impressions

• The CHPT Translate tool appears basic, resembling older neural machine translation interfaces with limited functionality.

• Users can choose from predefined prompts such as making translations sound more fluent or business formal, which raises questions about their logical application.

Critique of Translation Prompts

• The prompt "translate this and make it sound more fluent" is critiqued for lacking clarity regarding what it compares against.

• Discussion highlights confusion around the purpose of prompts; they seem illogical when applied directly to initial translations rather than post-editing.

Audience Reactions and Feedback

• Poll results show mixed reactions towards CHPT Translate: 45.6% find it mediocre, while 28% see it as interesting and 26% consider it significant.

• Comments from industry professionals suggest that the interface may have been hastily developed, indicating potential issues with its core product viability.

Service Now's AI Launch and Market Reactions

Google's Acceleration in AI

• Steven Holmes from Service Now views the launch as a response to Google's rapid advancements in AI, indicating a competitive landscape where OpenAI must adapt.

• The decision to focus on translation as a standalone offering highlights significant demand for AI-driven translation services.

Monetization Strategies

• The need for increased user engagement is emphasized, suggesting that more visits and clicks are essential for revenue generation.

• The introduction of a dedicated translation service indicates strategic moves by companies to enhance their monetization efforts.

Recent Mergers and Acquisitions

Italian Acquisition: Agletch 14

• Agletch 14, an Italian LSI, acquired Centrolingua SRL, marking a small-scale acquisition involving minimal asset transfer including one employee and some translation resources.

US Acquisition: Cineverse Corp.

• Cineverse Corp. has acquired Giant Worldwide, focusing on media services and localization to enhance its platform called Match Point.

• Match Point aims for end-to-end automation in digital entertainment delivery through integration with Giant's technical services.

Financial Insights

• Cineverse expects Giant Worldwide to contribute $15-$17 million in pro forma revenue for fiscal year 2027, indicating strong profitability potential.

Funding Activity Overview

Cynthia's Series E Funding

• Cynthia raised $200 million in Series E funding at a valuation of $4 billion, showcasing significant investor confidence.

UK Support for Tech Growth

• The funding round received notable attention from UK politicians, highlighting national support for tech innovation within the country.

Key Investors Involved

• Google Ventures led the funding round with participation from Nvidia’s N Ventures among other investors, reflecting strong backing from established players in the tech industry.

Discussion on AI Avatars and Language Integration

Overview of AI Avatar Technology

• The discussion begins with the potential of creating avatars from approximately two minutes of video, allowing for quick personalization.

• There is skepticism about the market size for companies aiming to change workforce training through this technology, despite a significant valuation of $4 billion.

Market Viability Concerns

• The speaker expresses doubts regarding the abstract nature of the use case and questions whether language capabilities are truly central to their mission.

• Initially perceived as language-centric, the company's vision has expanded into broader areas that may dilute its original focus on language.

Importance of Language in Global Enterprises

• A counterpoint is made emphasizing that while language may not be front and center, it remains essential for scaling and serving global enterprises effectively.

• Companies with multilingual workforces require comprehensive language support; limited offerings would not meet enterprise needs.

Future Engagement and Insights

• The speakers express interest in further engagement with the company to understand how they integrate language into their broader mission.

• Mentioned is a liquidity initiative for early employees, indicating financial strategies involving NASDAQ for secondary sales.

Funding Updates: Deep Graham's Series C Round

Financial Overview

• Deep Graham raised $130 million in a Series C funding round, achieving a valuation of $1.3 billion, bringing total funds raised to $215 million.

Company Focus Areas

• The company specializes in live multilingual speech-to-text services and voice AI applications across various languages.

Use Cases Highlighted

• Key applications include live customer support, conferencing assistance, clinical documentation, trading communications, and compliance monitoring—areas where accuracy is critical due to sensitive data involved.

Potential Expansion into Conversational Translation

• Discussion includes potential future developments in live conversational translation as an extension of their current capabilities.

11 Labs: A Potential AI Giant?

Overview of 11 Labs' Valuation and Revenue

• The Financial Times reported that 11 Labs is in early discussions to raise funds at a potential valuation of $11 billion, indicating significant interest and investment in the AI sector.

• The company has reported an annual recurring revenue (ARR) of $330 million, showcasing its financial performance and market position.

• If successful in raising funds at this valuation, 11 Labs would become the most valuable AI startup in the UK, highlighting the competitive landscape among tech startups.

The Role of London as an AI Hub

• The discussion emphasizes London's status as a central hub for AI development in Europe, suggesting a vibrant ecosystem for innovation and investment.

• Comparisons are made with other companies like Cynthsia, which has a lower valuation, illustrating the varying scales within the industry.

Umplara's Unique Funding Approach

Company Profile and Funding Strategy

• Umplara filed an S1 with US regulators to raise $100K through a share sale; however, this is not a traditional IPO but rather a self-underwritten direct public offering.

• The funding strategy involves selling shares directly by the director to interested parties rather than going through banks or exchanges.

Insights on Market Dynamics

• This approach is described as "survival capital," indicating that Umplara needs immediate funding to continue operations while seeking product-market fit.

• The conversation highlights how accessible it is to file for small amounts like $100K in US capital markets compared to higher barriers faced by companies in London.

Challenges Facing IPO Activity in London

Regulatory Environment and Market Conditions

• There are concerns about declining IPO activity in London due to high regulatory barriers; efforts are being made to lower these barriers.

• Rachel Reeves' initiatives aim to revitalize London's IPO market by making it easier for companies to go public, contrasting with more flexible practices seen in US markets.