La grande illusion
A diary that was more than four years in the making, from the “wake-up call” of 24 June 2016 to a solitary New Year’s Eve 2020. This is what Michel Barnier, the former negotiator of the European Union for Brexit negotiations with the United Kingdom, gives us in this book, which was published on 5 May 2021.
As is so often the case with a divorce, the narrative is shot through with sadness, but also tensions and confusion at dealing with a succession of British negotiators with profiles as varied as they were colourful and whose intentions were frequently obscure. Two central questions that still remain unanswered are the backdrop to the action: what were the British thinking of when they threw themselves into the Brexit adventure? And have they any more idea now than they did five years ago about what this project means for their country?
This Brexit diary is also a story of minor treachery and attempts to short-circuit the official negotiation channels, a practice that did not even spare Berlaymont, the headquarters of the European Commission, when Martin Selmayr still roamed the corridors of power, never afraid to “add fuel to the fire” by deliberately bringing pressure to bear on the negotiator (our translation throughout).
By talking about his part in this slice of history, Barnier obviously also tells us about himself, his values, his vision of the European Union, which cannot solve all the problems of the European people, but also his vision of France, now that he has returned to national politics and hopes to help, or even lead, his family of Republicans in the forthcoming presidential elections, even though he does not expressly state this intention in his book. We are already familiar with Barnier’s passion for hiking in the mountains, reflected in this book with memories of the Olympic Games of Albertville in 1992. The mountains, moreover, inspired many of the welcome gifts he gave to his opposite numbers. On 19 June 2017, for instance, he gave a walking stick to David Davis, the first British negotiator, who gave Barnier a book about Annapurna in return.
Courtesy, however, is often merely a facade and, with the exception of a few individuals such as Olly Robins, Theresa May’s negotiator, Barnier was not in all cases able to develop positive feelings about his interlocutors, including the aforementioned David Davis, whom he described as a warm man but a touch temperamental, and who seemed to be more concerned with himself and with the talks. The EU negotiator seems to have struggled on occasion to read Steven Barclay and Dominic Raab. His harshest judgement, however, is reserved for Jeremy Hunt, the British Foreign Secretary at the time, who we get the impression may have somewhat annoyed Barnier by accusing the EU of being to blame for the no deal outcome. As for David Frost, the final negotiator, Barnier appears to have failed to establish particularly cordial relations with this man who, he explains, constantly attempted to circumnavigate him, even up to 24 December 2020, when the two sides finally concluded an historic deal on their future relationship, once the final agreement on fisheries have been reached.
But for the entire length of the book, Barnier never waivers in his respect and even admiration for Theresa May, who had to suffer months of low blows from her own party and implement a decision to leave the EU that she had voted against. For the “great illusion” of the title of this work – the name of a film by Jean Renoir, and also of an essay by Norman Angell, as Barnier explains the beginning of the book – is ultimately a unique witness statement, on almost a day-by-day basis, of the complete and utter unpreparedness of the British to leave the EU and a constant improvisation act, set against the backdrop of the personal political ambitions of certain individuals, such as the future Prime Minister, Boris Johnson, or Jeremy Corbyn, the Labour leader at the time. This improvisation was accompanied by manoeuvres of less than total propriety, such as 23 December, when the British negotiators sent the Commission a “text full of traps, false compromises and backtracking” at two in the morning to try to discreetly bag some last-minute concessions, Barnier recounts.
This ongoing chaos on the other side of the English Channel, a permanent theme of this diary, simply reinforces the impression of apparent stability given by the European negotiator and thus the cohesion of the EU as a bloc. From June 2016 until early 2021, when the adventure came to an end for him, Barnier consistently stressed three major principles: preserving peace in Northern Ireland, defending the single market and maintaining the unity of Europeans in the response to the questions Brexit has raised concerning European integration. This triptych was painstakingly defended by the European negotiator’s team, almost exclusively made up of women, with the Commission’s Director General for Trade, Sabine Weyand, now deputy head of the cabinet of Ursula von der Leyen, Stéphanie Riso, the former head of cabinet of Jean-Claude Juncker, Clara Martinez, and the negotiator Paulina Dejmeck Hack.
And as if his mission has not yet finished, Barnier calls upon the EU not to let its guard down in its dealings with a United Kingdom which has not yet laid a certain “filibustering” to rest and which “will try to get into the single market by the back door” in an attempt to “mitigate the consequences of Brexit that it caused”. (Solenn Paulic)
Michel Barnier. La grande illusion – Journal secret du Brexit (2016-2020) (available in French only). Gallimard. ISBN: 978-2-072- 88001-8. 544 pages. €23,00
Les organisations de défense face aux défis de l’intelligence artificielle
Although by no means exhaustive, as the author freely confesses, the work is no less rich, tackling as it does all aspects of artificial intelligence (AI), on the basis of its current developments and applications, without leaving out a discussion of its hypothetical future potential. In it, the Belgian political scientist Alain De Neve deals with the technological dimensions but also, to a considerable extent, the political, social, organisational, strategic, legal and ethical considerations.
As the author points out, despite recent progress, artificial intelligence is still a capacity that stems entirely from its programming, in other words it has no genuine autonomy, but with an automation potential that may bring about risks. The computational power currently being developed, at exaflop level (i.e. 1018 operations per second) has the potential to resolve extremely complex problems in a timeframe that far outstrips human capability, in terms both of carrying out the operations and verifying the process and the result achieved. However, algorithms are susceptible to errors and the automated forms of learning that are progressively associated with them can also embed bias and incorrect data. Although we are still a long way from the AIpocalypse described by science fiction writers, there are many risks that will need to be taken into account for the future. These concern principally cyberspace and our societies, well beyond the military dimension or cyber-terrorism. Another threat is that of a new race to arms, on the basis of competition between China and the USA.
Central to this development is data, “which is, at this moment in time, the main challenge of the competitiveness of AI systems across the world”, according to the author, who adds that “without data, AI cannot learn. And in order to learn, the quantity of data required is unlimited. In other words, whatever the level and quality of data they can obtain and control, players dominating the AI sector (…) will never be able to get enough” (our translation throughout). This is the case even though the relevance and reliability of data are largely informed by the mass collected. If, however, in a military context, “the profusion of data from the various dimensions of the operational area (air, land, sea, space and cyber-space) has certainly been able to represent a vital asset in seeking informational superiority over the enemy (…), it has very swiftly become evident that a greater flow of data gathered by the various sensors deployed could end up slowing down the decision-making process”, the author adds.
“It is currently still very difficult – and, in some cases, impossible – to say whether the systems using AI are drawing the right conclusions or even how they draw them. These systems often look like ‘black boxes’ to researchers and operators. Sometimes, algorithms produce ‘odd’ results, resolve problems by using the wrong method or an illogical one, or even cheat”, De Neve reveals, stressing the “need to set validation and verification processes in place that are specific to this technology”. He also stresses the great vulnerability of these systems to data manipulation; a small change to a single pixel in an image can prompt an algorithm to identify a machine gun as a helicopter.
Although in general, “Europe is some way behind China and the United States on the techno-industrial level”, the EU and its member states have finally adopted strategies on AI. However, “the community of European users exports more data than it gathers” and “mastering the data is the weak link – or the missing link – of any scientific European undertaking in the field of AI”, De Neve notes, stressing that although the larger member states are investing significant sums of money in research into these technologies (1.5 billion for France, 3 billion for Germany between now and 2025), initiatives are mainly national, with cooperation constantly coming up against industrial turf wars, like the recent Franco-German controversies over the joint future fighter plane project.
AI pits our defence companies and institutions against “new challenges which, far from being insurmountable, call for a debate to be clearly defined, not simply concerning tomorrow’s combat methods, but, even more so, concerning the adaptation of our decision-making mechanisms in the event of crisis”, De Neve concludes, stressing that “this point becomes even more urgent in the context of nuclear deterrence and the temptation to entrust AI systems with certain activation keys in nuclear exchange scenarios”, more specifically in the context of a second-strike capability. (Olivier Jehin)
Alain De Neve. Les organisations de défense face aux défis de l’intelligence artificielle (available in French only). Institut royal supérieur de défense. Sécurité & Stratégie 146. ISSN 0770-9005. 125 pages. This study can be downloaded free of charge from the IRSD website (http://www.defence-institute.be )
Le prophète et la pandémie
Gilles Kepel guides us through the soap opera of the year 2020, that was marked, amongst other things, by the pandemic, the attacks in Nice and Vienna, the brutal murder of Samuel Paty, the continuation of military operations in Syria and Libya, the toppling of alliances, the rise and rise of Erdogan who became, over the months, the principal agent of regional destabilisation in the Mediterranean and the Near East and, of course, the departure from office of Donald Trump. The story is told breathlessly, like a radio drama in which events come thick and fast, with the odd flashback to previous episodes and, inevitably, a few repetitions.
The political scientist devotes many pages to what is effectively a reconstitution of the strategic backdrop, the so-called Abraham Accords, allowing the state of Israel to be recognised by the United Arab Emirates, Bahrain and then Sudan, with the support of Egypt, the United States and, more discreetly, Saudi Arabia. According to Kepel, this agreement, which was announced at the White House on 13 August, “makes the Jewish state the hinge of the tripartite axis of the Muslim Brothers and the Chi’ites, which grew up over the course of the second decade of the 21st century between Turkey, Qatar and Iran, bringing together the politico-military objectives, beyond the once-fashionable Arabism, of three states that speak respectively Turkish, Arabic and Persian”.
“In the coming clash between, on the one hand, this axis and, on the other, the Abraham Accords, alliances with shifting borders and informal mutual obligations, the global powers – the United States, the EU, Russia and China – are more or less trying to keep everybody happy on the basis of their own interests, avoiding an unequivocal engagement that could have the consequence of a global conflict”, Kepel stresses, adding that “as usual, the European Union is manifesting its lack of power and its spinelessness: under the Presidency of the German Chancellor during the second half of 2020, it strictly rationed its empathy for its own member states, Greece and Cyprus, in the face of military incursions by Ankara and– like NATO – expressed no support for France in the incident of 10 June between the frigate Courbet, which was attempting to inspect the Turkish cargo ship Cerkin in the framework of the NATO ‘Sea Guardian’ operation, as it was suspected of carrying weapons to the Libyan port of Misrata, and the German-built frigate Oruç Reis”. Without attaching too much importance to the Franco-Turkish naval incident, it is true that NATO’s silence can only be seen by Erdogan as a free pass. More scandalous, in a different way, is the lack of solidarity expressed towards Greece and Cyprus by certain member states of the EU, principally Germany and Italy. Even more scandalous is without doubt the press statement made by the President of the Italian Council, Giuseppe Conte, the day after the European Council of 1 October, explaining that Germany and Italy “are only attempting to impose a constructive dialogue approach (…) so as to bring about a de-escalation. This involves recognising a strategic role for Turkey. It has an important role to play in the eastern Mediterranean, in the Middle East, the Balkans and also in Libya”. However, as Kepel rightly points out, “the Turkish government cannot afford a stand-off with a Europe that is united and determined, which it needs far more than the other way round”. (OJ)
Gilles Kepel. Le prophète et la pandémie – Du Moyen-Orient au jihadisme d’atmosphère (available in French only). Gallimard. ISBN: 978-2-0729-2312-8. 324 pages. €20,00