grand plans (not always well thought through), not enough by way of inclusivity, the prioritisation of industry, and a disturbing lack of attention to the risks posed by AI
(10 points quickly jotted down based on a copy of the recently-released Draft National Telecom Policy shared with me; the draft policy is available online too)
1. Although one wouldn't expect this to be in the policy, it's remarkable that one rarely hears talk of basic literacy without which access to digital anything is difficult. There's also no specific mention of enhancing access to the digital world, so to speak, for people with disabilities.
2. The policy seems to cobble together terms which aren't necessarily meaningful in conjunction with the words near it but which sound as though they're at the cutting edge of technology; the oft referenced ‘quantum’ seems to be a favourite.
e.g. the aim to "Strengthen the security by adopting quantum resistant cryptography" — what's being strengthened exactly and how?
3. Surely, having everyone complain is not how the non-functionality of systems should ideally be assessed?
re “Formulate stringent QoS parameters for network availability (wireless/wireline), Broadband (wireline) service and customer services, focusing on Quality of Experience (QoE) and utilizing crowd-sourced data” reiterated with “Review QoS benchmarks emphasizing Quality of Experience (QoE) and real user data”
Also, there are such systems in place to detect network outages. They require a significant number of complaints to be made to be meaningful — those complaints are not always made and there's no mention of a plan to modify user behaviour in this regard.
4. That said, as optimistic and upbeat as the document is, it doesn't simply build castles in the air. Talk of satellite networks is, e.g., backed up by talk of securing orbital slots.
5. It's not at all clear why the goal to “Encourage low-latency communications for industrial applications” should be restricted to industrial applications in general rather than to sectors critical to the public.
6. If one were to “Create an innovation centric section 8 company” per a stated goal for 2030 — never mind why… What would such a company actually do?
7. Other goals are astonishing. An oblique reference to the ISM band in the goals for 2030 says: “Make spectrum available for R&D with minimal compliance requirements and in alignment with global best practices and standards” — Shouldn't this already have been in place and, if not, why can't it be expedited?!
Also, it appears that industry will take precedence in the ISM band (although the draft doesn't say which frequencies it has in mind when it says e.g. “Make available spectrum for Captive Non-Public Network (CNPNs) and private 5G for industrial use cases”).
8. The aim to “Develop a framework for accepting IPRs as collateral by financial institutions to support innovation financing” sounds interesting though, of course, it needs to be complemented with a standardised framework for IP valuation which we don't really have.
9. “Establish a biometric based identification for all telecom users to ensure privacy and protection” — no explanation as to why biometrics are needed to ensure privacy & protection provided, what form biometric-based identification would take (unless it's merely Aadhaar unnamed), and how it would be implemented.
10. The enthusiasm for AI seems unbridled and there's no talk of human oversight. AI is repeatedly mentioned and some plans such as that to “Collaborate/facilitate relevant ministries/departments for a resilient Public Protection and Disaster Relief (PPDR) network with harmonized spectrum plans, AI-powered predictive analytics, and IoT-enabled systems for resource mobilization” are particularly discomfiting given that predictive AI is notoriously unreliable and the plan doesn't so much as mention the possibility of human intervention to mitigate errors.