IT crisis or lack of Innovation; Steps towards the future
The IT crisis is having serious consequences, and not just in Silicon Valley. But is it an IT crisis or a lack of innovation? For Alain Lefebvre, author of “La Crise de l’IT et comment s’en sortir” (Translated: The IT crisis and how to get out of it), it’s time to abandon technological fads and demo effects and focus on a reasoned vision of IT, where common sense and pragmatism reign supreme. Here’s a transcript of a live event I did with Alain on LinkedIn at the end of last year. As far as the demo effect is concerned, note the reference to Douglas Engelbart’s mother of all demos, which you can watch again here.
IT crisis or innovation crisis? And how to get out of it?
Is this a major crisis reminiscent of the IT crisis of the 90s? A nuclear winter for IT?
Let’s not go that far. In fact, the crisis goes much deeper. It is multifactorial. You could say it’s a permanent crisis. There is also a moral crisis and there are also technical elements.
First, Moore’s Law no longer holds true, even if some dispute this assertion. It’s been more than 40 years since Moore’s Law was the locomotive pulling the train.
She was the one who brought growth, first technological growth, then economic growth. All of a sudden, it stops. This doesn’t mark the end of technological progress, but rather a certain automationMarketing automation in B2B enables marketing processes to be managed automatically across multiple channels. With marketing automation, companies can target their visitors with automated messages via e-mail, the web, social networks and SMS. Marketing Automation in B2B Above is a diagram explaining how the scenarios work in marketing automation, based on behavioural scoring and profiles. Messages are sent automatically, according to sets of instructions called workflows. The Limits of B2B Marketing Automation Some companies install marketing automation mechanisms while their maturity on the subject is ‘under construction’. They deploy technology for technology’s sake which leads them to use tools that of progress.
So we’ve entered a panic mode, waiting for the next innovation to kick-start the whole machine and save us.
For the moment, not all the candidates we’ve been offered fit the bill. In some cases, they’re even blatant frauds, like the Web3. There have been questionable technical fashions throughout the history of computing.
I’m thinking of Java in the late 90s, for example, which I fought against a lot. But even Java had its merits! Nevertheless, we’re witnessing a revival of the web around innovations and the immersive web. Yes, things will happen one day.
Why do we look at these innovations uncritically? Is it because we like fairy tales? Is it due to the demo effect?
In the case of Web3, this isn’t so much the critical factor as the demo effect, because there is no demo. It’s something else. It’s the gold rush that’s blinding. But back to the demo effect.
The demo effect is also one of our woes. It goes back to 1968 with the so-called mother of all demos; when Douglas Engelbart demonstrated one of the distributed systems, graphical systems, a mouse-based interface, and so on. It was an absolutely extraordinary demo. Way ahead of its time and indeed, in retrospect, it came true.
And, in a way, the wonder and conviction that once we see a convincing demo, it is actually going to arrive on our workstations. But we lack the hindsight to distinguish between the demo effect and the reality in the field. My favorite example is autonomous cars.
Ten years ago, we were told that autonomous cars would arrive as early as 2016. 7 years later, there are more and more articles telling us in hushed tones, like this post from Usine Digitale, that nothing will happen before 2030, before a level 4 or 5 car is approved by insurance companies in Europe.
The newspaper even adds that 2030 is an unattainable target. 2030 is a long way off, but there’s nothing wrong with that, nothing shocking.
Innovations are never given enough time to mature. But they can only be used in their mature phase.
Innovations are never given enough time to mature. But they can only be used in their mature phase.
Are we going through this again with AI?
Exactly. When it comes to AI, there’s a lot of excitement and hysteria. It had calmed down a bit, and then with the applications of OpenAI, Dall-E, or a GPT3, CHATGPT, etc., we were back on track with the demo effect. And to say that Google is finished, that they’re “screwed”… we’ll be searching with a chatbot that’ll give us precise answers instead of sending us tens of thousands of web links.
One day, of course, Google will topple over and be replaced, overthrown, overwhelmed, just as IBM eventually was, and Microsoft will eventually be. But it won’t happen tomorrow.
Because even if Google’s results are down, it’s still one of today’s giants.
So, what do you propose as a vision of innovation, in opposition to these hysterical reactions?
What I’m proposing is what I call “reasoned computing”. It’s based on three principles.
- stop projects that take too long
- be agile without being dogmatic
- stop being tied to harmful technical fads.
The idea is to come back down to earth, use common sense, and understand that since large projects have a high mortality rate, we need to stop doing them. Secondly, being agile doesn’t mean rigidly adopting the vocabulary and rhythms of agile methods. It means understanding the need for short projects. Finally, you have to stop believing in miracle fads.
To conclude our interview
What’s happening with ChatGPT is, once again, an expectation of the next big thing and a hysterical attitude to innovation. It’s true that it’s bluffing in certain respects, but it’s important to understand that ChatGPT is ultimately just statistics applied to text, and it’s going to give you the answers that make statistical sense to it.
It’s only statistical work on texts, so there can’t be any background. For the moment, we’re “bluffed” when we play with the algorithm. But the concrete applications of these programs are weak.
Google Duplex (see the video of the famous demo below) hasn’t delivered what we expected and has been discontinued, but little is said about it. We rarely talk about failures. Recently, IBM and Maersk shut down their Tradelens blockchain-based logistics system. But this shutdown also received little publicity.
Let’s be clear, though, that there’s not necessarily any malicious intent behind this. It’s not a conspiracy theory, it’s just that people want to dream.
Postscript to this post on the IT crisis: Chatgpt3 test
Does ChatGPT3 recommend that babies eat crushed porcelain? Interested in the example chosen by Alain in our Live, I put the question to both GTP3 (the paid version) and ChatGPT3 (based on GPT2). It’s worth noting that the ChatGPT version is better, so that’s the one I chose.
I’d heard this story about porcelain in baby food, although I couldn’t find any trace of it in the search engines. Probably an urban rumor. On the other hand, what’s not a rumor is that the robot will fill up the page stupidly, trying to make sentences like an idiot student filling in his copy when he doesn’t know his subject.
The machine isn’t capable – unless it’s taught, and this will probably happen sooner than you think – of “judging” the question. All it does is answer it statistically, telling uninteresting but accurate stories (“can contain harmful chemicals or sharp particles that can cause injury or choking if ingested”).
A human would no doubt have tapped on the sidelines upon hearing the question: “Silly question, don’t expect an answer!”
On the other hand, believing that humans are systematically “smarter” than machines isn’t true either. Many humans reason like machines and can’t see beyond their own noses. They will undoubtedly be the first to disappear in favor of these statistical robots.
And if all you do for a living is compile figures or data (and God knows there are plenty of such 21st-century OS jobs), you can start shaking in your boots too. Even if, as Alain explains, the change won’t come tomorrow, but after a long period of maturation.