DeepSeek: Consumers win as tech innovation moves to next stage

Taken at its word, the China-based DeepSeek is delivering the breakthrough technology at a fraction of the cost and energy consumption. Picture: AFP

Security issues aside, the emergence of DeepSeek in the generative AI sector is undeniably good news for the global economy and consumers.

Taken at its word, the China-based company is delivering the breakthrough technology at a fraction of the cost and energy consumption.

If AI is the ground-breaking technology it is assumed to be, then what’s not to like about that?

If you are Jensen Huang at Nvidia, obviously your 90 per cent gross profit margins are at risk and that explains why his stock price was hammered early this week, before recovering some of the gains on Wednesday.

Investors were dealt lessons.

It was instructive that Apple, a perceived AI laggard, saw its stock price rise on the same day Nvidia fell 17 per cent.

This was a classic new technology reaction, with producers ­enjoying all the early gains, explaining Nvidia’s 55 times price-earnings ratio valuation, but in time users and consumers win as competition and commoditisation threaten producer margins.

At the outset it should be noted there are many sceptics who don’t accept the DeepSeek claims, and there is government concern expressed at a Chinese company’s ability to efficiently hoover up data around the world.

TikTok is banned from the app market in the US, but DeepSeek can access the data.

Four years ago, Chinese telecommunications equipment provider Huawei was banned from the 5G mobile phone rollout in Australia, which in turn sparked a local merger between TPG and Vodafone.

Regulations can have unintended consequences, need questioning and updating because they cannot keep pace with technology.

A spokesman for the ACCC said: “We welcome competition and do consider there would be value in different choices and different price points being available for consumers, subject to any considerations by the government of national security issues.”

Competition drives more innovation and lower prices, which benefits the whole market.

The early AI movers have operated under an infrastructure maximisation strategy: more Nvidia chips, more data centres and more energy. Investors have blindly followed the mantra. It would help if the dial could be wound back – less energy and less data.

Nvidia’s growth was fuelled by the belief that its chips alone were indispensable for training large model language models. That is now being questioned, and better still by a company opening its models for all to see.

Laggards such as Apple and Meta might be able to participate in the AI market, which was previously considered out of reach, not to mention start-ups that are just entering the game.

Doors may open.

The AI boom, like other technological revolutions, was built on a belief that everything will change, rather than actual evidence of reliable change.

The key theory is known as the Jevons paradox, named after 19th century economist William Jevons, who noted about the coal market that improved efficiency lowers the relative cost and increases the demand. Now the world is moving to the next stage.