The first parts of this series explored the promise of generative AI along with the sizable technological, policy, and cultural hurdles facing adoption across communications carriers. While by no means exhaustive, several proven recommendations emerge from the early experiences of providers pioneering ambitious implementations amidst legacy environments:
1. Launching Generative AI "Centers of Excellence"
Leading carriers like AT&T, Disney and Samsung have launched dedicated generative AI leadership teams consisting of executives, data scientists, engineers, business partners and ethics advisors. These centralized groups act as hubs of expertise guiding education, governance, strategic partnerships and talent development while generating IP through cutting-edge research pilots. They crucially also maintain connections with frontline domain experts across customer service, network infrastructure etc. to ensure models adapt appropriately per use-case nuances. Formalizing such centers of excellence removes organizational siloes, encourages collaboration, and signals commitment from the outset to integrate generative AI responsibly.
2. Investing Into Cloud and Data Infrastructure Upgrades
Deutsche Telekom serves as an ambitious example of a carrier navigating deliberate, multi-year initiatives to migrate individual network domains to cloud-based platforms primed for AI integration, such as modern digital BSS/OSS systems and decentralized 5G edge networks. While the technology transformations span years, each incremental move enables onboarding real-time analytics and automation use cases on newly transitioned domains in parallel that provide quick generative AI wins to maintain momentum.
3. Developing "Guardrails by Design" Frameworks
Before models reach real-world testing beyond contained environments, instilling constraints and monitors to track issues like toxicity, bias and inaccuracies is vital. For example, Anthropic'sConstitutional AI approach offers a model "bill of rights" encoding safety, security and oversight at the algorithmic level itself vs. playing catch up later. Such proactive "Guardrails by Design" reduce risks as carriers deploy generative AI conversing directly with consumers at scale.
4. Nurturing Responsible Generative AI Startup EcosystemsThrough Venture Funds
Telcos like Orange and NTT run active generative AI venture funds that provide startups financial runway, commercialization launchpads via pilot testing bandwidth and high-visibility co-marketing opportunities. Beyond returns, such funds allow carriers to forge early partnerships in frontier tech areas ahead of natural commercial inflection points in coming years while proactively shaping ecosystem ethics standards.
5. Offering Reskilling Programs for Existing Employees
Software giant SAP demonstrates the efficacy of global reskilling initiatives like "AI Business School" teaching employees hands-on machine learning skills applied on company data to retain talent by unleashing their passions. Telecoms require similar large-scale internal talent transformation and change management drives to augment niche external generative AI recruiting. Upskilled domain experts adopting emerging techs propagate adoption further.
The Responsible Path Forward
Generative AI brings boundless opportunities from customer experience to infrastructure automation for telecoms. But prudent steps around modernizing technology while nurturing ethical culture and talent set the stage for smooth adoption at scale. Communications companies embracing principled progress over perfection can responsibly unleash generative AI’s near-term productivity benefits while shaping its responsible evolution through steadfast guardrails and governance. With pragmatic expectations, this enormously promising technology category holds potential to transform connectivity services over coming decades just as the internet did through the turn of the century.
Sources:
1.https://www.bcg.com/en-us/publications/2020/how-to-launch-ai-excellence-centers
2.https://www.lightreading.com/aiautomation/cloud-infrastructure-crucial-to-telco-ai-ambitions/d/d-id/761525
3. https://www.anthropic.com/blog/constitutional-ai
4.https://fortune.com/longform/ai-artificial-intelligence-ethics-big-tech/
5.https://about.att.com/innovation/whitepaper/our_principles_for_ethical_ai
