Don't forget to check out our other Electronic Design Quick Polls.
What you’ll learn
- What about mask mandates?
- Is it time to stop using 2FA short messaging service (SMS)?
- Is AI taking over the world?
According to Merriam-Webster, skepticism is “an attitude of doubt or a disposition to incredulity either in general or toward a particular object, the doctrine that true knowledge or knowledge in a particular area is uncertain, the method of suspended judgment, systematic doubt, or criticism characteristic of skeptics, doubt concerning basic religious principles (such as immortality, providence, and revelation).”
Skepticism can be a good and bad trait depending on how one follows up on any doubt regarding a particular topic. Three “facts” have been hot topics these days including “wearing masks does not prevent the spread of COVID-19,” “two-factor authentication (2FA) is no good,” and “ChatGPT will make programmers unnecessary.”
The challenge for a skeptic is that many are trying to push fear, uncertainty, and doubt (FUD) to make you want to believe in something that may not be true. In these cases, there may be a nugget of truth, but the main premise could be false or at least misleading.
In the first case, the Cochrane Review report is what pushed COVID to the forefront. However, before you take their “facts” as truth, you might want to check out responses such as “Masks Reduce the Risk of Spreading of COVID, Despite a Cochrane Review Saying They Don't” that states:
“An updated Cochrane Review published last week is the latest to suggest face masks don’t work in the community.
However, there are problems with the review’s methodology and its underpinning assumptions about transmission.”
The problem is actually quite nuanced and worth the trouble of understanding the issue. Unfortunately, many will take statements from one source at face value. Personally, my family still uses masks as my wife is also a microbiologist, over 65, and immunosuppressed, so it’s something we have more knowledge about than the average person.
The second case involves Twitter deciding that 2FA SMS will no longer be free to those who don’t pay a fee, with the justification that 2FA can be bypassed. The part about “in some cases …” or “better alternatives” seems to get lost in the shuffle.
This case tends to be less controversial than the COVID issue, since more of the experts agree that 2FA is useful, recommended, and that cyberattacks on 2FA SMS can be easily mitigated with education. 2FA also is a tool that embedded developers contend with when designing systems.
Finally, we have OpenAI’s ChatGPT which cuts closer to home for our readers (also see the ChatGPT Quick Poll at the end of the article). The plethora of articles about it and related software makes the other two issues pale in comparison. ChatGPT has been noted to pass law school exams. It can even write software programs.
ChatGPT and its ilk have great potential, but they are really in their infancy at this point. Like the saying, “garbage in, garbage out” (GIGO), one must look at how the machine-learning (ML) algorithms are built and trained.
Those unfamiliar with the state of artificial-intelligence research and products these days are easily misled, or they can make invalid assumptions and conclusions about the technology spreading FUD. It’s a topic that we’re examining more deeply at Electronic Design since ML/AI are tools and topics being utilized throughout the engineering community.
Another example is Github’s Copilot uses OpenAI’s Codex. Copilot has been trained using billions of lines of code that are public Github software projects. The projects span dozens of programming languages. It is designed to generate “the boilerplate and repetitive code patterns” that are common in these projects. Of course, it assumes you want to repeat these patterns and spending $10 to $20 per month is worthwhile. Splitting the good code from the bad may be yet another AI project.
Closer to home, confusion may still reign if one makes assumptions. Our article “The Right Development Tools Help Avoid Costly Delays” highlights the difference between free versus paid development tools. Unfortunately, some may assume that this means open-source vs. paid or open-source vs. closed-source tools.
It’s true that there are free open-source and closed-source tools—open source usually means free, at least in terms of access to, and distribution of, source code. The problem is more nuanced, though; you should check out the article for more details.
To wrap up, a good skeptic is one who looks to understand an issue as well as the sources of information. Likewise, changes over time can make a big difference in an issue or topic. Vaccinations have made a major impact on COVID, 2FA SMS remains a viable and useful feature, and ChatGPT can get some things right, but it also can be very wrong.
So, while I trust vaccines and masks at this point, I look at ChatGPT as an evolving science project that holds promise but isn’t a production tool for programmers. I will let you come to your own conclusions, but be skeptical.