pl | en

No. 229 June 2023


images „High Fidelity” | Piksel Studio

No 229

June 1, 2023

Or „High Fidelity’s” 19th Anniversary

OU HAVE PROBABLY NOTICED THAT - one of the hottest topics in recent months has not been the war in Ukraine, not inflation, and not even politics, but ChatGPT (Chat Generative Pre-training Transformer). It was launched in November 2022 by Open AI, which calls it its "child":

We’ve trained a model called ChatGPT which interacts in a conversational way. The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests., accessed: 26.04.2023.

The GPT-3.5 version available at that time stirred a lot of emotions and caused an outpouring of comments, analysis and various assessments. From technical, through economic, psychological, and sociological, to social, concerning equality, wages. And there is already a ready version of GPT-4, which is to beat the previous one. Despite such a variety of fields in which this topic is discussed, one thing resonates in all of them: the fear of "what if...". This 'if' is usually tinged with the fear of being replaced by a machine learning computer program.

Because this "intelligent" bot can do everything that does not require physical work: write a poem, paint a picture - this is how the cover of the latest issue of the "Polityka" weekly was created - will pass exams in virtually any subject, write a book, compose music, and all this in any style set by the user. That is genius. However, I have my own opinion about its genius and versatility, based on reading, conversations and experience. I think the fears are greatly exaggerated.

For it is the case that until such a program becomes self-aware, it will only be a compiler - brilliant, fast, and efficient, but nothing more. If we tell it to learn something based on social media, it will tell lot of stupid things and "lie" - I use it in quotation marks, because this will not be a conscious reaction, but a learned mechanism. ChatGPT does not distinguish information from opinion, it does not know jokes, it cannot properly react to irony. Most of all, however, it is not creative.

It sounds a bit like I'm rationalizing my own fear. And maybe it was like that in the beginning. But the more we know, the calmer I am. I believe that people whose work consists in producing valuable, qualitative "things" will get more work, not less. Only those who did not understand their work, who did it poorly or even did not know anything about it, may feel threatened. They will be replaced in no time.

⸜ The first ever „High Fidelity” cover, at the time it was still rectangular, № 12, April 2005

The problem, however, may be what will happen to the recipients of the "artifacts", i.e. texts, information, videos, images, etc., which will come from the bot. Most of them won't even notice the change. And if so probably for the better. Because instead of idiotic entries in social media, on blogs and even in newspapers and online and printed magazines, they will get well-formulated, legible messages adapted to their level. The fact that it will be untrue, stupid, and sometimes simply dangerous news - it will not reach them.

And I am not an elitist, but a pragmatist: in 2019, the PISA (Programme for International Student Assessment) analysis was published, a project conducted by the OECD (Organization for Economic Cooperation and Development), and its research showed that about 40% of Poles do not understand what they read, and another 30 have great problems with understanding written text. Which means that only 30% of people from our beautiful country can read texts quite efficiently (more → HERE). And this is not a result that differs dramatically from the results of studies from other parts of the world.

In the text entitled How to fight secondary illiteracy? posted on the website of Polish Radio IV, we read:

Dr. Paweł Boguszewski, a neurobiologist from the Institute of Experimental Biology, pointed out that for many reading comprehension is a skill that they cannot use in everyday life. – Analysis of a literary text can be a learned activity. We may know how to do it, but not necessarily know how to use it to our benefit. Research shows that an increasing number of Poles have problems with reading loan agreements, filing taxes reports, reading more complex texts, he enumerated., accessed: 26.04.2023.

So how can we not say that ChatGPT, much more efficient at creating false information than people, will not seduce these 70%? This is actually certain.

⸜ In February 2008, for the only time in our history, we had two covers, displayed randomly, „High Fidelity” № 46

But to the point. The biggest problem for me is the lack of distinction between facts and opinions. Although the term "alternative facts" has been coined and some journalists assume that there are "ours" and "their" facts, the true ones are always on our side, but this is logical nonsense and idiocy, because the facts are one and indivisible. The rest are not.

Such a mix of information and opinions has been visible for a long time in the news published by a large number of audio magazines. When you read in the news that "award-winning loudspeakers use unique technologies that translate into low distortion and incredibly natural sound", then know that you are reading a text sent to the editorial office, written by people from PR. Text untouched by the editor's hand.

While we only have information here that new loudspeakers have been developed - the rest is either an opinion, such as "translating into (...) incredibly natural sound", or difficult to verify "throw-ins", such as "they use unique technologies" and "resulting in low distortion”. For it to make actual sense, the reader should be informed which is a concrete information and which is the manufacturer’s an assessment (distributor). Can a bot do that? I don't think we'll even know how we've been manipulated by it because it'll put all the required quotation marks, "–ed." and "by" in its text, but not necessarily where they should actually be used. We will get a falsehood that perfectly imitates the truth.

But there's nothing to be afraid of. It was like that in the past, it is happening now and it will happen that there are people who want to cheat and others the former want to cheat. And ChatGPT is actually a fantastic tool that will help you speed up work and relieve us of automated, simple activities. It can also be helpful in running your own audio magazine, such as "High Fidelity". And although for now we do not plan to use a bot, despite the fact that you can find many such information aggregators on the web pretending to be magazines. , I would like give some advice to those who would like to create tests, columns, etc. at low cost, in a word - to have their own audio magazine "by ChatGPT". Maybe they'll come in handy.

⸜ The first ever square cover for „High Fidelity”, „High Fidelity” № 52, August 2008

First of all you need to have an idea. It is not enough to copy others, because other bots will do it better. For me, "High Fidelity" was from the beginning a magazine, a monthly. That's what I had in my head and that's how I thought about it. Therefore, it has a specific date for the publication of a new issue, it is the first day of a given month, as well as a division into topics such as in a printed magazine. Since April 2005, each new issue has a cover - a graphic on the right side of the table of contents. Anyway, the table of contents is also an idea from the world of print. Your ideas may be different.

Secondly, you need to know your craft and constantly improve it. For me, it was important that I could combine technical knowledge, both the one acquired at school and at work, with the language, i.e. with what I learned at university. It may look different in your country, many things can be learned while working. However, you have to start somewhere, you need some initial capital.

It might seem that ChatGPT will help with this, but everything indicates that its "handling" will require even more skills than before. What the bot writes will be deceptively similar to the truth. Just as mixing truth and lies is the best way to brainwash, there is a danger that such "washing" can be done by our own, grown on our own breast, bot.

There's something else too - we must love what we do. The Internet, apart from the dangers, has also brought us an increase in trust in authentic people. That is, people who believe in what they do and say. Thanks to this, of course, many individuals emerged with completely twisted personalities, often even terrifying, but also others whose self-confidence gives positive results, in a word, is a positive factor. From what I've seen, the Open AI bot can't handle it at all, because it's not creative. But that will probably change as it learns from us, until most people decide that in what it does it is real.

⸜ The last „High Fidelity” cover using the old layout, № 62, June 2009

Still, this is the most important thing, we have to be honest about all this. Responsibility for the written word is one of the foundations of journalism. Who will be responsible for what the bot writes? – Publisher? Journalist? Or maybe the producer? After all, no one has any idea what is going on "inside" the ChatGPT decision system, we only know the result. Perhaps the way out of this situation will be to treat irresponsible texts, dangerous texts, hateful texts or mud-throwing texts, stupid texts and texts containing false information as if they were created by bots? This would solve several problems, because we would know in advance that there is some intention behind it, someone wants to manipulate us, or on the contrary, there is no mind behind it, zero living cells, which will relieve us even more from the need to read. We can then delete them immediately.

Taking all this into account, I believe that a new, really wonderful time lays ahead for "High Fidelity". For nineteen years, we have grown and matured together with you, our readers. And if so, then - I know it for sure - you are sensitive to the lack of professionalism, lack of commitment, carefree writing and insincerity, otherwise you would have abandoned us after the first text.

We make mistakes,, we stumble, imperfections happen to us - we did all that. However, you can be sure of one thing - we do it for you and with you in our minds. And we have a lot of fun, because it's an important part of our lives. We can make a decent living from it, which allows us to focus on "High Fidelity" to the fullest. Thank you for still being with us and we ask you for even more years!

Chief editor

About Us

We cooperate


Our reviewers regularly contribute to  “Enjoy the”, “”“”  and “Hi-Fi Choice & Home Cinema. Edycja Polska” .

"High Fidelity" is a monthly magazine dedicated to high quality sound. It has been published since May 1st, 2004. Up until October 2008, the magazine was called "High Fidelity OnLine", but since November 2008 it has been registered under the new title.

"High Fidelity" is an online magazine, i.e. it is only published on the web. For the last few years it has been published both in Polish and in English. Thanks to our English section, the magazine has now a worldwide reach - statistics show that we have readers from almost every country in the world.

Once a year, we prepare a printed edition of one of reviews published online. This unique, limited collector's edition is given to the visitors of the Audio Show in Warsaw, Poland, held in November of each year.

For years, "High Fidelity" has been cooperating with other audio magazines, including “Enjoy the” and “” in the U.S. and “”  in Germany. Our reviews have also been published by “”.

You can contact any of our contributors by clicking his email address on our CONTACT  page.

positive-feedback linia hifistatement linia Net Audio

Audio Video show

Vinyl Club AC Records
Audio Video show