Like many people I’ve been following closely both the hype and reality of ChatGPT. At the same time, I’ve been learning a lot about different forms of machine learning, and how they can be used to enhance computing, especially as computers grow in power. I also have been using computers for more then a quarter century, so I have some ideas on where machine learning could be useful for every day uses, while other cases were not so useful.
Things I Don’t Think We’ll See in the Future
ChatGPT is not going to replace writers, artists, or secretaries except for the most basic tasks
ChatGPT will not ever write quality news articles, press releases, or publications.
Natural language searches will not become the norm, as it’s a lot of typing or speaking and prone to mistakes, however machine learning will continue to be applied to both search terms and results to get more useful results
Things I Think We’ll See in the Future
Machine learning will be applied to people’s personal computer file system to better flag mistakes in documents, like in Microsoft Word. If for example, you regularly type out a press release or a report a certain way in Word, and something is different format-wise or stylistically in your current version, your Word processor would flag if not automatically fix it.
Machine learning could automatically generate templates based on previously saved documents on your computer, allowing you just to update and fill in the details of the document.
Machine learning would be used for resizing and colorizing photos in Gimp and Adobe Photoshop, automatically tracing edges, vectorizing and detecting words.
Machine learning would make the creation of graphics more automated, by creating sensible styles, and anticipating your next move.
Command line code would be far better automated, with much better tab competition both based on the commands others have used, and what you have previously run
Basically, any process you run on your computer would have much more tab completion, with the computer automatically predicting your likely next move, helping to speed up processing, as the computer could start working on the likely next step
I don’t anticipate the centralized machine learning model, with vast databases getting that much play. Internet access can be funky, and people are often hesitant to share data. It’s risky to be too reliant on other people’s servers. But I do think machine learning is going to only grow in importance on desktop computers, with more and more predictions made locally to assist users in getting tasks done quicker on their computers.
On Friday, a group of Extinction Rebellion activists occupied a Tesla store in central Milan. Activists chained themselves to the cars’ tyres, and others glued themselves to the windows along with the slogans “Make millionaires pay again” and “Ecology for all, no ecofascism”.
Analysts are openly wondering if Musk is causing lasting damage to a brand he has made synonymous with electric cars and, by extension, liberal aspirations to tackle climate change.
The SSD drive is for purposes of booting so much faster then a conventional drive — taking my laptop about 4 seconds to boot into Linux with the XFCE window manager compared to 30 plus seconds with the traditional drive.
Applications load somewhat faster, but the real speed difference is when you are preforming operations that are highly disk intensive, such as loading a large database or processing many digital elevation models at once. Obviously, CPU intensive applications aren’t improved by SSD, so if something involves more processing then loading, the benefits will be smaller.
All together though, I’m much happier with the speed of my machine with the SSD drive installed. It’s well worth the $150 to be able to start and shutdown the machine quickly, and have more access to my files. Plus, I still have the old hard drive in the DVD drive bay, so I have ample space on my laptop for years to come.
