It is now safe to establish that every Dutch person comes into contact with artificial intelligence on a daily basis. A weather report on the iPhone, a search on Google, an automatic tag on Instagram, a recommendation on AliExpress. Most of those applications come from the US or China. That the Netherlands (and the rest of Europe) is lagging far behind in its own artificial intelligence (AI) is an urgent problem.
The Netherlands is in fact becoming a data colony of the US and China. That is also the tone in the Strategic Action Plan for Artificial Intelligence that State Secretary Mona Keijzer (Economic Affairs and Climate, CDA) presented on behalf of the government on Tuesday. But the plans to turn the tide are not convincing. The Netherlands is relatively late with its strategy compared to other countries. The reactions to the content are also critical. Coalition partner D66 wants Keijzer to tighten up the plan.
Professors and experts from the sector mainly see missed opportunities. The scope: it is too little, too late. The critics are right. The plan mainly brings together previously committed budgets. The 'billion euro' in the press release mainly consists of promises about possible funds to be tapped and previously promised funds. For comparison: the Chinese cities of Tianjin and Shanghai both made more than 12 billion euros available for AI last year. The cabinet formulates three 'tracks for policy'. The three are 'Exploiting opportunities, creating the right conditions, and strengthening foundations'. Those are points with which nobody will disagree. But large, politically sensitive questions remain unanswered. Should there be restrictions on takeovers of AI companies by Chinese or American competitors? Should a new joint academic top center be established to stop the exodus of AI talent? How is there more capital for AI start-ups? How are rights safeguarded when AI makes decisions about people?
Concrete political ideas such as a plan from VVD and D66 for a special algorithm authority to supervise new applications of AI are apparently in the refrigerator. Also a plan to work on secure online identities, important for the management of data that artificial intelligence feeds on, is not addressed in the strategy.
The strategy gives a confused, superficial and strongly compartmentalized impression. It is more about polishing the status quo than the concrete step forward that is needed so much. Bright spots are also there. It is good that science is being encouraged to formulate joint research agendas around artificial intelligence. Plans to stimulate the sharing of data between Dutch companies sound promising.
On Wednesday, the World Economic Forum think tank came up with a ranking of the most competitive economies in which the Netherlands emerged as the best European country. The Netherlands has a lot to catch up with. But with this strategy, the government is missing an opportunity to be at the forefront of Europe and to set the agenda. The European Commission wants to quickly update its existing AI strategy. European Commissioner Margrethe Vestager was asked by the European Parliament this week about her plan to come up with a fully revised European AI strategy within a hundred days. Isn't that too ambitious, the parliament wanted to know. “Yes, that is very ambitious,” she answered. Mona Keijzer will not get that question soon.
source https://www.nrc.nl/nieuws/2019/10/11/nationale-ai-strategy-goes-trap-not-reduce-a3976384
Tags: AI, Nederland, opinie