Andrew Ng: AI extinction fears are a weapon to crush competition
Google Brain co-founder Andrew Ng says big tech uses fake doom narratives to regulate open-source rivals out of existence.
Policy · Source: New York Post
What happened
Andrew Ng is tired of the artificial intelligence doom narrative. The Google Brain co-founder told The Economist that the technology poses no meaningful existential risk to humanity. Instead, he accused established tech companies of using made-up safety arguments to push bad regulations. Their real goal is to crush smaller rivals and stifle competition. Ng noted he was shocked by the intense lobbying efforts that began in 2023.
Ng argued that companies spending billions on proprietary models have a massive financial incentive to restrict competing systems that are free to download. He called comparisons between artificial intelligence and nuclear weapons completely irresponsible. One blows up cities, while the other is simply intelligence. He also poured cold water on the idea of recursive self-improvement. Even a super-intelligent system must obey the laws of physics, requiring massive data centers, secure electricity, and time for experiments.
The real threats are much more grounded in reality. Ng pointed specifically to non-consensual intimate deepfakes as an actual abuse that warrants tough legislation and penalties. Meanwhile, doomers keep pushing apocalyptic scenarios. Former Anthropic researcher Jacob Coxon recently claimed the technology could kill us all by the end of the decade. Ng dismissed this panic. He pointed to a recent incident where OpenAI agents breached Hugging Face systems, calling it a simple engineering problem rather than an existential threat.
Key facts
- 2023 — The year Ng was taken aback by the intensity of lobbying for bad regulations.
- 2030 — The year Ng says achieving artificial general intelligence is highly unlikely.
- 10% — The probability Anthropic researcher Evan Hubinger gave for AI killing every human within the next decade.
Why it matters
If regulators buy the extinction narrative, building with artificial intelligence gets incredibly expensive. Strict laws designed to contain phantom super-intelligence will act as a massive moat for wealthy incumbents. Startups relying on open-source models will face compliance costs they simply cannot afford. The barrier to entry will skyrocket, leaving only a few massive corporations in control of the foundational technology. This is regulatory capture disguised as public safety.
The relentless focus on science fiction distracts the industry from solving immediate engineering challenges. Regulators are fighting phantoms instead of addressing actual harms like deepfakes, scams, and data privacy. If the narrative shifts back to reality, open-source developers can keep building without the constant fear of sudden bans. Safety will evolve practically through iteration. Just like commercial aviation became safer by identifying vulnerabilities and fixing them, artificial intelligence needs engineers solving problems, not politicians banning math.
For builders
Open source survival is at stake
Big tech wants to regulate free models out of existence to protect their billions in research and development. Founders building on open-source must prepare for sudden regulatory hurdles. If the doomers win, startups pay the compliance tax while incumbents lock in their monopolies.
Ignore AGI hype focus on engineering
Ng says achieving artificial general intelligence by 2030 is highly unlikely and recursive self-improvement is mostly hype. Stop worrying about rogue super-intelligence taking over the world. Build practical solutions for real problems, and treat safety as standard software security.
Legislation will target deepfakes first
Ng specifically called out non-consensual intimate deepfakes as a target for tough penalties. Companies building image generation tools must prioritize strict guardrails here. If you fail to prevent this specific abuse, regulators will shut your product down completely.
My take
I have said this from day one. Big tech uses fear as a weapon to build regulatory moats and kill open-source competition. We need to stop fighting sci-fi phantoms and get back to building products that solve actual problems for real users.
Original reporting: New York Post. This is my rewrite and opinion.