AI: Risk Vs Reality

Friends, let me clear the AI noise today in plain text.

PART 1: THE RISKS BEING RAISED — WHO, WHEN, WHY, WHAT

WHO IS RAISING IT:

- OpenAI CEO Sam Altman

- 128 states at UN Convention on Certain Conventional Weapons in Geneva

- US and China ahead of AI safety talks mid-September

WHEN:

- This week, Sep 15-16, while Brent oil holds near $108 and Fed decision is due today at 2pm ET. Tech shares fell in Tokyo and Seoul, SoftBank down.

WHAT WAS SAID:

- Sam Altman warned of 10 percent risk AI could cause human extinction by end of this decade if ungoverned. He has warned before, but timing matters now.

- In Geneva on Saturday, 128 states agreed first non-binding text on autonomous weapons, so-called killer robots. US and Russia opposed binding law, want national guidelines only. Rights group Stop Killer Robots said definition was watered down.

- US House Speaker Mike Johnson said Tuesday: There cannot be a moratorium on AI development without losing competitive edge to China. They can self-regulate.

WHY NOW:

Because AI can now do most laptop tasks — coding, designing, writing, hacking. That accelerates three real risks: bioweapons design, automated cyberattacks, and autonomous killing machines. That is why regulators are moving now, and why markets reacted.

My take: Self-regulation did not keep Strait of Hormuz open. It will not keep killer robots governed.

PART 2: THE POSITIVE SIDE MEDIA WILL NOT ADVERTISE

Media sells fear because fear clicks. History tells a different story.

HUMANS HAVE ALWAYS BEEN SCARED OF NEW TOOLS:

- PENS AND WRITING: Socrates warned writing would destroy memory.

- PRINTING PRESS 1440s: Kings and churches said it would spread heresy. It spread literacy.

- TYPEWRITER 1870s: Said it would destroy handwriting and put clerks out of work. It created modern offices and jobs for women.

- TELEPHONE 1876: Said it would destroy face-to-face talk. It connected families across continents.

- COMPUTERS 1970s-80s: Said it would make everyone jobless. It created the entire digital economy.

- INTERNET 1990s: Said it was for hackers and porn, end of truth. It is how you are reading this.

Every time the pattern is the same: FEAR, THEN RULES, THEN ADOPTION, THEN ABUNDANCE.

AI IS THE SAME, BUT MORE POWERFUL. SO WE NEED GOVERNANCE, NOT PANIC.

WHAT AI IS ALREADY DOING POSITIVELY TODAY:

1. FOR FARMERS IN AFRICA: AI weather models plus soil analysis raising yields 20 to 30 percent without extra fertilizer.

2. FOR HEALTH: AI reading X-rays in South Sudan and India, detecting TB and cancer where no doctor exists.

3. FOR EDUCATION: A child in Juba can now get a personal tutor for math, English, coding — free — that we never had before.

4. FOR BUILDERS AND BUSINESS: You can now build a business, design a logo, write code, translate Chinese supplier contracts — without hiring a team. Portable income.

5. FOR GOVERNANCE: AI tracking corruption in procurement, tracking ships around Hormuz and Bab el-Mandeb when three routes fail at once.

CONCLUSION: GOVERN THE TOOL, DO NOT FEAR THE TOOL

The danger is not the tool. The danger is who governs the tool, and for whom.

We should NOT ban AI. We should NOT let 5 companies self-regulate humanity's future.

We need just and equitable global security governance — same as Defence Minister Dong Jun called for at Xiangshan Forum in Beijing this week — but for code. Binding rules for autonomous weapons. Shared audits for frontier models. Investment so Global South builds with AI, not just fears it.

Typewriters did not end writing. Printers did not end truth. Computers did not end work.

AI will not end humanity — unless we let it be governed by fear and greed instead of people.

Skill before scale. Systems before shortcuts. Integrity before income.

Build with it. Govern it. Don't fear it.

John Monyjok Maluth

Emerging Africa-China Scholar