Unpredictable 2021 - FrenchWeb.fr

At the end or beginning of each year, I usually ask myself from a distance about these exercises, often vain, consisting in trying to predict what will happen in the next 12 months. Predictions have been even more off the mark than usual in 2020.

Few people had imagined how Covid-19 would turn the year upside down, upset our relationship to time, to others, to space, to risk, to death, to the delicate balance between our personal freedoms and common good. A difficult year which also allowed us to struggle, to help our relatives and friends from a distance and also to strengthen ties despite the distances. Of course, we also think of the dedication of health professionals, as well as those who ensured the operation of essential activities such as the food trade.

Simulation and determinism

During confinement, we have almost become laboratory rats, more or less locked up in a large or small home. The situation is testing our emotional resilience to endure these distanced and rather formatted human relationships.

This indirectly revives the debate on the deterministic side or not of our world. What is deterministic and what is not? Is the world we live in just a vast simulation? Does he leave a part to chance or not?

In Do We Live in a Simulation? Chances Are about 50–50 by Anil Ananthaswamy, Scientific American, October 2020, the question is posed of discovering the clues as to whether we live in a simulation or not. It makes a useful detour through quantum physics and the fledgling techniques of simulating this physics using quantum simulators. It is a separate class of quantum computers capable of simulating the interaction between elementary particles such as atoms. If we can exploit them today to simulate a molecule of a few atoms, we are very far from being able to do so on a simple polypeptide, even less for a protein, and even less on a cell and a fortiori, on a human body. So, all over the Universe! This tends to invalidate the simulation hypothesis.

Elon Musk thinks he can “matrixize” the world with his startup Neuralink, which would connect our brains to AIs. In short, if we were already in a simulation, it would create a simulation within the simulation. Could this meta-simulation be infinitely recursive? And preserving human vision with nearly 90 million pixels per eye? There is a sizing problem in this case!

The simulation argument is a complex epistemological question. What is a simulation? The fact that the physics of this world is governed by fundamental laws that are all quite simple (gravity, fundamental forces, …) does it make it a simulation orchestrated by a machine or something intelligent? Or is it a simple self-regulating natural balance, mega-scale swarm physics?

For those who have come across it, the simulation argument clashes with notions of complexity theories that are shattered. There continues to be debate about whether or not certain classes of problems will be treatable with quantum computers. We just stop at the so-called NP problems whose solution is verifiable in polynomial time. But it is not the most complex class of problems there is. What is the complexity class of the hypothetical simulator of the world? What would be its technical characteristics? What laws of physics would have to be circumvented or violated for it to be feasible? The practical dimension of these basely material things seems to escape theorists of the simulated world.

In any case, this feeling of living in a simulation has been exacerbated by the experience of covid-19 which places us in unprecedented situations, in particular in an isolation that is difficult to bear. But this situation is not so extraordinary for those who know or have known wars.

On the other hand, the determinism of the forecasts has been put to the test. In a digital field alone, I don't remember seeing anyone predicting the acquisition of Xylinx by AMD, Slack by Salesforce and arm by Nvidia. Nor does anyone predict the coming advent of videoconferencing that would work the first time in all circumstances. I have the impression that this feat will see the light of day long after the advent of scalable quantum computers, expected depending on the case in 10 to 40 years….

With that, I've been giving you a lot of fun with quantum technologies in 2020 and it's not going to stop anytime soon. For lack of time, I did not spend the year commenting on a whole bunch of other areas. This post gives me the opportunity to catch up a little bit.

Covid after effect

The year 2020 was difficult to predict because of what the covid-19 pandemic triggered. 2021 is not easier to anticipate. We think we see the end of the tunnel, but the deadline remains unclear.

During the first confinement, it was customary to imagine a world after different from the world before. The deconfined summer was an opportunity to verify that the world after was very similar to the world before, and sometimes worse. We dream of flying far away, of having parties with friends, of rediscovering a world of encounters, serendipity and discovery. Or a pretty normal life for those who don't have access to all those luxuries.

We will perhaps expand our relational and emotional field, rediscover human relations IRL, the thirst for truth and contact, but at the same time appreciate teleworking when we are employed. And also perhaps better choose the people with whom you want to work and live. In terms of technological uses, it goes without saying that the pandemic has reinforced and will continue to reinforce the deployment of teleworking tools and videoconferencing of all kinds. It could also help popularize augmented and virtual reality systems, just to escape in an even more realistic way without moving (with VR). But are we then going to walk in the street in such a way as to no longer be recognized at all, the top of the face being covered by an augmented reality system and the bottom by a mask?

In a very down-to-earth way and in his abundant predictions for 2021, Fred Cavazza foresees, among other things, the development of local online commerce. Why not. In the USA, C-NET adapts to the situation and helps you choose your yoga mat, your side sleeping mattress and your pillow. The world of tech has evolved a lot! In Forbes, Sachin H. Jain anticipates the emergence of the issue of mental health at the end of the pandemic. What will be the real weight of the trauma suffered by some and will the return to normal be enough to heal the wounds?

The other delayed impact of the pandemic is that which weighs unevenly on sectors of the economy. The stock market was not affected when logically it should have been. But since prices anticipate the future, which cannot be better than today, it is perhaps understandable that the stock market is bullish instead of bearish. But that's just dust piled up under the rug. Until when ? This will become a hot topic in 2021 and 2022.

And then, we will finally be able to see the last James Bond. However, “No time to die” was a very timely imperative during the pandemic! We replaced that with great series like “The Queen Gambit” and “For All Mankind”.

Deployment of 5G

For years, equipment manufacturers and telecom operators have been telling us about 5G. She finally arrives. The announcement of Free in December launched hostilities well. The question is: where, what for and in what order? Specialists announce it primarily in business applications, in industry 4.0, in connected objects, in connected cities and health, and even in autonomous vehicles. If those are potential markets for 5G, my logical little finger tells me that's not where it's going to start.

I'm not going to give you a rundown on the technological richness of 5G, which is a simplified acronym covering all the diversity of the telecom world, all imaginable frequency bands, and all uses ranging from very low-speed machines to machines at very high mobile speed from human to human. But it was not the telecom players who invented the Uber, Airbnb, Whatsapp, Instagram and other TikTok applications!

My little finger tells me that the start of 5G will first go through the adoption by the general public of 5G smartphones which are now flooding the market, especially with the latest iPhone 12 which are all 5G. The cheapest 5G smartphone is 300€ at Oppo. They will soon be even cheaper. It was predictable at the end of 2019 with the announcement of Qualcomm's Snapdragon 765G processor. What about apps? The first application of 5G in our cities and countryside will be to create 4G that works! Who hasn't cursed with a deplorable flow in a simple café in town? 5G technologies will improve the speeds and quality of service of our mobile connections. Our videoconferences will work better when we are on the move (!). In short, in most places, including in town, 5G will not be 2 GBits/s but simply a correct speed greater than 10 Mbits/s and which we can therefore use more comfortably than before. This can help facilitate teleworking and nomadic work.

On this, consumer killer apps may be launched in the tradition of Instagram and other Tik-Toks that have enhanced 4G. They may be mixed with virtual or augmented reality. We do not know anything. Just as no one could foresee the mass adoption of Pokemon Go in 2016. With that, attempts to elevate our consumer experiences will surely be launched.

On the smartphone side, manufacturers will continue to create bizarre formats to create buzz and eventual growth. Like the LG Wing, this rotating double screen that unfolds to create a T-format screen, horizontal at the top and square at the bottom, which allows you to hold your phone vertically when it is in horizontal mode. It opens with the thumb and closes with two fingers. This follows the Samsung Galaxy Fold V1 and V2, and other Huawei with foldable screens.

None of these formats will succeed. They complicate the life of developers too much, are not standardized, and unnecessarily weigh down smartphones. However, these will continue to improve in terms of photo and video functions, driving the nails in the coffin of the camera market which has been under attack for more than ten years.

Artificial intelligence

AI is obviously ubiquitous in the technological sphere. An AI did this one day, that another day. It gets boring because it's so repetitive. Especially since the AIs in question are almost always the same method: the use of data from the past to vaguely predict the future, and, for the moment, without reasoning. Today's AI is confined to the antiquated intelligence of a narrow-angle rearview mirror. The goal of many researchers: to create AIs that need less training data. In short, who will be able to reproduce even better the human biases that often come from a lack of training data!

In Forbes, Rob Toews predicts that a political deep fake will spread widely in the US and spread confusion and misinformation. Yet they don't need AI to make it happen, with their incumbent president capable of all of that, and even worse. The author predicts that publications by researchers on federated learning will increase, which is easy to anticipate with a trend that started only three years ago. This technique created by Google aims to better distribute the training of neural networks between cloud and smartphones, thereby protecting your personal data used for the training in question.

It also predicts that a major AI chipset startup will be acquired by a major semiconductor player for more than $2B. Well, a bit like Habana, acquired by Intel at the end of 2019… for $2B. He thus predicts rather deterministic events: the fact that a language processing model will exceed one trillion parameters (in the American sense of the term), which is nothing extraordinary since the record is already 600 billion parameters with Google's G-shard. In any case, we will hear more and more about AI that would titillate general artificial intelligence with transformers, compositionality, and other areas that are not easy to grasp, which above all improve the language processing capabilities of AI. So, as long as you make predictions, you might as well stick to what's already been launched. Moreover, the bricks of AI will continue to interfere in all business software. It is becoming distressingly commonplace. AI is a kind of logical continuation of the software developments of the past decades.

For PwC, the most important application of AI is in cybersecurity and the fight against fraud. InformationWeek anticipates that generative (aka “deep fake”) networks will have applications in media. At ThinkML, we are predicted even more RPA, this branch of AI that automates processes in companies. All believe that language processing will continue to progress in 2021.

Many predict even more interesting uses of AI in healthcare. Contrary to appearances, the AI ​​bricks were very useful during the covid-19 pandemic even if it was not in spectacular use cases. This will continue to progress in 2021, in particular with the deployment of AI-assisted medical imaging applications as well as in drug discovery methods.

Quantum technologies

With and without confinement, 2020 gave me the opportunity to immerse myself in science as I had never done before, and also to discover or better know a great number of researchers in the field. This was reflected in the latest version of my ebook on the subject (684 pages).

The year was quite prolific, with announcements from Honeywell, from IonQ, from the Chinese with their boson sampling experiments, with the publication of IBM's ambitious roadmap, with the arrival of Amazon in quantum hardware (which predicts mountains and wonders in 2021), and more recently with a quantum simulation carried out with 196 cold atoms by an international team including that of Antoine Browaeys (IOGS and startup Pasqal). The significance of the latter is major.

The next stage of all this will be to obtain a “quantum advantage” (performance of a quantum computer superior to that of the best classical supercomputers to deal with a given problem) on useful and truly programmable machines. This could very well happen in 2021.

The Quantum Daily team predicts that there will be staffing difficulties in quantum, which is not surprising, as funding arrives more quickly in the field than the training of talent which takes several years. It also anticipates an increase in public and private funding. After a record year in 2020 with more than $1B in startup funding.

I also found a large package of forecasts on quantum technologies at Thinkers360, an intermediation platform for luxury thinkers and influencers. The latter would have done better to choose another field because they often tell nonsense on the subject, in particular Chuck Brooks, Frank Feather, Jai Singh Arun, Stewart Skomra and Chan Naseeb. No time to fact-check everything, it's a Danaïdes barrel! The profession of consultant who fact-checks other consultants does not seem very promising to me in 2021.

The proliferation of R&D in quantum technologies is in any case taking a turn that recalls that of the beginnings of the history of computing in the 1940s and 1950s. The end or slowing down of the effects of the law of Moore in silicon re-opens the game. She shuffles the cards and extends the fields of what is possible next. While it is the determinism of Moore's law that has kept optical computing, superconducting computing and reversible/adiabatic computing off the path of mainstream technology for several decades. The attenuation of this determinism generates a greater opportunity to innovate. This also makes it dizzy because there is so much scientific and technological news in the field. In other cases, the determinism is very present, such as that which makes it possible to predict that Apple will launch new iPhones around September/October 2021.

And then, what becomes of the Arlesian of the French national quantum plan? This should finally be announced in 2021. This is a paradoxical and relatively rare situation in politics. Indeed, this plan has been ready since July 2020. The announcement was to be made during the autumn by the President of the Republic. It was delayed because of the reconfinement launched at the end of October 2020. We are therefore waiting for the next deconfinement. But meanwhile, part of the operationalization of the plan is underway. The SGPI will audition in January the candidates for the coveted position of the national coordinator of the plan. Calls for research projects are in preparation at the Ministry of Research. But the quantum ecosystem is getting impatient because at the same time, other countries have been much quicker to relax, especially the Germans who announced a €2 billion plan in June 2020. Plan which is now looking for its projects! But rest assured, the French plan is not a new “calculation plan” which would decide in a top-down way how the industry should structure itself.

Cryptocurrencies and blockchain

If I happen to express myself on these two subjects where I am to be placed in the camp of skeptics, it is only in this exercise of forecasts.

I am always taken aback by cryptocurrencies and in particular by the Bitcoin yoyo. The semantic and practical confusion is always maintained between investment currency or refuge (much like gold), currency of exchange and means of payment (for shopping). The price of Bitcoin was around $26K at the end of December 2020 and a CitiBank analyst predicted that it will reach $318K by the end of 2021. This does not at all argue for making it a real currency.

And then, it's not just Bitcoin in life. There is also StableCoin, Ripple, LiteCoin and Ethereum. What a mess! How to find it? It's a bit like there are 30 PC or smartphone operating systems out there! This is at best a symptom of a still very immature market. Facebook's Libra has become Diem. Never mind ! Consumer benefits? Enclose us in one more walled garden? We still do not know what this can be used for, at least in well-banked countries such as Europe.

And when we see that some central banks are launching or are going to launch their own cryptocurrency, it makes you wonder insofar as, in the original politico-technological discourse, bitcoins were supposed to keep us away from these irresponsible central banks!

What about Blockchain applications? Exponential forecasts are also going well on their subject. Even if it does not seem to be the great excitement in the business projects by exploiting.

The Saxo Bank, which always gives in the outrageous in its 2021 forecasts, predicts that the blockchain will make it possible to stop fake news! Ah-ah-ah. For 2019, the bank's analysis predicted that Apple would acquire Tesla, that Donald Trump would fire the Fed boss (he's one of the few he didn't fire…) and that Jeremy Corbyn would become Her Majesty's Prime Minister. Missed !

Digital sovereignty

2020 was the year of renewed interest on the part of States, especially European ones, for the notion of digital sovereignty. Under the leadership of Thierry Breton, the European Union has followed suit. The pandemic has accentuated the trait. I can tell you bluntly that this will continue in 2021! But we will often take a wrong turn, acting on the effects and not on the causes of the digital non-sovereignty that is rampant in Europe. The codes of American dominance are still not well understood. We want to protect our internal markets when the priority should be to attack external markets. The Americans are crushing us thanks to the size of their homogeneous domestic market and their economies of scale, which are then replicated on a planetary scale, at least in the West.

Fortunately, some French startups have understood this. The French entrepreneurial ecosystem has held up rather well in 2020, supported in particular by numerous support plans launched by the government. Even if it is a perverse indicator, many beautiful startups managed to finance their international growth at a quite respectable level in 2020 (notably Alma, AB Tasty, Aircall, Contentsquare, Dataiku, DNA Script, Lydia, Mirakl, MWM , Owkin, Saagie).

The State has just announced support for sovereign 5G intended for private companies, in the form of a collaborative research project led by the IRT b<>com located in its establishment in Lannion in Brittany , with the expected hiring of 90 people. Sovereignty seems to have become the magic name for obtaining public funding!

2021 will also be the year of the renewal of antitrust actions against the major American digital companies, even Chinese with Alibaba, which is being questioned at home. It will probably be a decade and it will end in blood sausage, as was the case for IBM in 1982 and for Microsoft in 2002. We are talking about the dismantling of the giants (Google and Facebook), and we will ultimately have a little regulation lukewarm.

Be careful also not to lose our sovereignty in other ways, in particular by the vulnerability of our information systems shamelessly attacked by state hackers, especially Russians, as with the recent SunBurst attack. This will also concern the USA, whose Biden administration is considered less friendly, to put it mildly, than that of Trump, vis-à-vis Putin's Russia.

Science and society

One of the topics that concerns me at the moment is the increasingly tenuous link between science and society. The Covid-19 pandemic and the associated conspiracy on hydroxychloroquine, vaccines, masks and 5G has highlighted this societal gulf and how the difficulty of understanding a complex world is turning into conspiracy. This phenomenon is very well explained here. What's more, France turns out to be the country where skepticism is the most developed with regard to science and technology.

Is it the fault of the elites, education and the obscurantism of certain politicians which was well highlighted in 2020? Here too, we must beware of looking for scapegoats and rather dig into systemic explanations, precisely. This is all the more so since the phenomenon affects all social classes. I was even able to observe it with unexpected audiences, including higher scientific training. We live in a schizophrenic world where there has never been so much knowledge on the internet and so much ignorance in people's heads. As time is limited and the brain can be manipulated by their emotions, they easily deviate from reason. If it was only for love! But, no, that's for everything! It is a phenomenon that the politicians, especially the extreme camps, know how to exploit very well.

Complex subjects such as AI and quantum, but also everything related to energy, give rise to their share of misunderstandings, opening the door to fanciful interpretations. Hence the need for continuous educational and precision efforts to try to calm things down. Etienne Klein regrets that engineers are not visible enough in the media. It's true. However, I will extend the remark to scientists in general. How to highlight them more? Train them to speak? Help them simplify without betraying? There is room for innovation from this point of view in mainstream media. I wish it more than I anticipate it for 2021! But do we know?

Seers

I always end this assortment of forecasts with those of the seers who always make people smile and are an exaggerated reflection of the previous debate. They are the mirror of a society driven by the irrational and also by lies. Many people accept them, sometimes knowingly. How to take all this seriously? It's still confusing.

To begin with, a certain Alice Bell, who comes to us from Brexit-affected Britain, predicts that 2021 will be better than 2020! Yippee. Just to be precise, June and July will be good months. Well yes, it's summer and the start of the holidays. Incidentally, we can hope, the vaccinations will have made good progress and it will be the return of exotic trips. On the other hand, she recommends that we remain open and flexible. Obviously, it is not necessary to follow the movement of planets and constellations (which, let us remember, do not correspond to very specific cosmic objects…) to know that!

Then, 2021 is a year placed under the sign of the number 5. Well yes, 2+2+1=5! Magic ! For Eva Cardini, this number 5 can both protect and attack us. We narrowly escaped the quantum number! Making predictions by indicating that a coin could fall heads but also possibly tails, that's how not to crash. It must be said that precision is not the strong point of clairvoyants and clairvoyants in general. Example: “In the meantime, there will be significant difficulties in France and in the countries of Europe”! What would be extraordinary? A year without difficulties!

And the top: “Some small businesses in serious difficulty will unfortunately file for bankruptcy”. For information, there were 52,000 business failures in 2019, in France. For the international situation, she foresees difficulties in all the countries examined. In general, clairvoyants and other astrologers are always rather pessimistic. Conversely, fortune-tellers convey more dreams! Everyone has their own marketing positioning in the chain of emotions!

Nicolas Duquerroy gives in the cafe of political and economic trade as many seers and in a text peppered with spelling mistakes. It predicts economic difficulties, business bankruptcies, a decline in GDP, more unemployment, a reconfinement at the start of the year, demonstrations (which would almost be a nice sign of a return to normal…), a better form for the National Rally (the seers are rarely centrist) and above all, original, of the “animalist party”. On the international menu, health problems for Trump – who is already mentally ill – and an attack in Germany, which happens from time to time.

If you want to know more and benefit from personalized predictions, you can choose from this database of clairvoyants and clairvoyants. Please note that the photos and first names of the clairvoyants are not contractual and the clairvoyances are broadcast by telephone on premium rate numbers. Selling dreams or nightmares is a business!

Next

This year, as I announced in January 2020, you won't get my usual copious CES report. However, I will speak on the subject in various virtual events, often with my sidekick Fanny Bouton. And there, without having to move since CES 2021 will be entirely virtual. I will miss this adrenaline rush of the start of the calendar year!

I am also currently working on updating my massive ebook on AI, the last edition of which dates back to the end of 2019, with a related project on energy still under way. Quantum news will keep me quite busy. And I have a few other educational content projects in the pipeline to brighten up your curious neurons!

With that, I wish you a happy new year 2021. Make it nice for yourself and for others! A lot of things are up to us!

The expert

Olivier Ezratty is a consultant and author, creator of Free Opinions, his blog on deep tech (artificial intelligence, quantum computing, medtech, etc.) and innovation (entrepreneurship , public policies…). Olivier is an expert for FrenchWeb which takes over the publications of his blog.

The ExpertsThe Experts are independent contributors to FrenchWeb.fr. To communicate on FrenchWeb or the Journal des RH, become a partner, click here.The latest articles by Les Experts (see all)TagsAB Tasty Aircall Alma Apple ContentSquare Dataiku DNA Script Donald Trump Facebook google IBM Jeremy Corbyn Lydia microsoft Mirakl MWM OWKIN Saagie Tesla Thierry Breton