My latest video looks at ChatGPT alternatives that can be operated on personal computers, including PCs and Macs.
I first look at Nvidia’s Chat with RTX, a tool enabling users to run a ChatGPT-like chatbot locally. Chat with RTX only works with Nvidia’s newer 30 or 40 series GPUs, which could be a limitation for some users. I tested it on a Lenovo Legion 5 Pro (affiliate link) that had an RTX 4060 GPU on board. Disclosure: the laptop is on loan from Lenovo.
I then tried GPT4All, an alternative open-source large language model client that offers similar functionality to Chat with RTX but without the need for high-end GPU hardware. Like Chat with RTX, GPT4all is user-friendly, requiring minimal setup and no advanced developer tools. GPT4All is compatible with various operating systems, including Macs, Linux, and Windows, broadening its accessibility. However, for optimal performance, 16 GB of system RAM is recommended especially on Windows.
In testing these platforms, I observed that while these AI models are capable, they are not nearly as good as ChatGPT. My test involved having the AI’s summarize one of my prior video transcripts for a blog post. I found that they more often than not got the context of the video wrong and even made stuff up rather than adhering to the facts in the source text it was summarizing.
But this does show how fast AI technology is moving from large data centers into something that can be run locally on a laptop. I was particularly impressed with how fast and responsive GPT4All was on my M2 Macbook Air as compared to a Lenovo Thinkbook running with a 13th generation Intel processor.
Both chat clients allow the user to choose from a number of different large language models. Although I only looked at three of those models in the video, there are many more offered as a free download to explore. These models are being updated all the time so I’m sure we’ll see some rapid improvements as the year progresses.
I’ve been around for awhile in the tech media space so I’m always weary when the next new “shiny object” emerges on the scene. Google Glass, VR, crypto and NFTs were mega hyped by influencers only to fall way short when it came to mass consumer adoption.
Over the last several months the chattering influencer class has shifted focus almost entirely to artificial intelligence (AI) driven by the very rapid advancements in Large Language Model (LLM) chatbots like ChatGPT. I haven’t heard a peep about NFTs in months!
I approached this new technology with a healthy degree of skepticism. While it certainly has a “gee whiz” factor to it could it actually have some real utility in my day-to-day life?
I decided to pony up the $20 monthly subscription fee for ChatGPT Plus to see if it could save me some time and make my workflow more efficient. And surprisingly – it did. You can learn more in my latest video.
I’ve been using ChatGPT to help write these blog posts based on the transcripts of my YouTube videos for the last month or two. Last week ChatGPT became even more useful through the introduction of plugins that allow ChatGPT to perform tasks that go beyond its pre-existing knowledge cutoff of September, 2021.
One of the plugins I’ve been using is VoxScript, which can pull down full video transcripts from YouTube which the ChatGPT can use to produce summaries for this blog and my email newsletter.
Here’s how it works: I provide ChatGPT with the URL of my YouTube video and ask it to write a summary in the first person in a journalistic, neutral language style. ChatGPT uses VoxScript to pull down the full transcript from the video and starts writing the summary. The result is usually a well-written summary that captures the key points of the video, saving me about 30 minutes to an hour of writing time.
The AI does an impressive job of interpreting the automatically generated YouTube transcripts, even correcting inaccuracies and presenting the information in a coherent manner.
Of course, it’s not perfect, and I do have to tweak some parts to ensure it aligns with my voice and style. But overall, it can generate anywhere from 75-90% of the post depending on what the topic is. This post, for example, needed a little more work done to it by yours truly but the framework it provided was a great time saver.
As AI technology continues to evolve, I’m excited to see how it can further enhance productivity and efficiency in various fields. And AI is more than just chatbots. For example Tesla’s full self driving system is an artificial intelligence neural network running locally on their cars trained to drive a car.
As always, I’m interested in hearing about your experiences with AI. If you’ve found a practical use for AI that has improved your workflow definitely head over to YouTube and share your experiences in the comments section of the video.
In my latest video I dive into the world of AI-powered music discovery with the Plex Amp player and its new “Sonic Sage” feature. Sonic Sage uses ChatGPT to deliver playlist recommendations.
Here’s how it works: Sonic Sage interfaces with OpenAI’s GPT model. To get it running, you’ll need an API key from the OpenAI platform. There is a small cost for using this key but I’ve found it to be minimal. So far I’ve only racked up about 5 cents of cost for well over 20 queries.
Once you’ve enabled Sonic Sage, it lives right inside the search icon on your Plex Amp app. ChatGPT uses your queries to generate music recommendations. You can ask it for anything, from general genres to very specific prompts. For example, you could ask for “high energy, lesser-known female rockers from the last 20 years”, and Sonic Sage will whip up a playlist to match.
The AI’s recommendations are based on how you word your prompts. While it’s not perfect at always getting things right, it does a pretty solid job of delivering great music to match what you’re looking for. The only drawback I’ve noticed so far is that these AI-generated playlists can’t be saved, but I’m sure this could change in the future.
This feature works best with a very large personal library or with Tidal, a subscription music service that integrates with Plex and Plex Amp. Tidal costs $8.99 a month if you subscribe through Plex and delivers all of its music as CD quality lossless FLAC audio. I covered the Tidal integration in a previous video.
In my view, Sonic Sage adds an interesting new dimension to Plex Amp’s already awesome music discovery capabilities.
Yes, as a lifelong nerd I of course had an Apple Newton back in 1994 when I was a senior in high school. Check out this photo I took of my Powerbook and Newton together likely in 1994 or 1995:
The Newton had a companion app for backing up and synchronizing called the “Newton Connection Kit” or NCK. Apple being Apple this was an optional software package that did not come bundled with the device.
One of the neat things about the Newton Connection Kit software was that it allowed data from most of the core Newton applications (calendar, notes, contacts, etc) to be edited on the host Mac or PC. It stored everything in a single file that I have preserved all of these years.
Every so often I boot up Basilisk II on my modern Mac and tool around with some of the old Mac apps I enjoyed as a kid. (This process, by the way, is a heck of a lot easier now with the awesome Infinite Mac website that’ll load things up in a web browser with zero configuration).
During one of these tinkering sessions I stumbled across the file and remembered how the NCK used to work. So I headed over to the Macintosh Repository and found a working copy of NCK, installed it, and then discovered that my old backup file wouldn’t open.
This was due to the file losing its Mac file type and creator information. Unlike the PC that uses a file extension (like .DOC for a Word document) the classic Macs stored the originating application inside the file itself. Oftentimes transferring the file from a Mac to a PC like I must have done years ago stripped the Newton backup file of this information. Because of this the NCK software couldn’t see or open the backup.
I searched extensively over the last couple of weeks looking for the four letter code needed to get this file open to no avail. So I asked ChatGPT 4:
Its answer of “pkgb” was unfortunately not the right one. It was close though – pkgb is the file type for Newton’s equivalent of apps that the NCK software would also install back in the day. Many times people would email these apps to each other and found the NCK software didn’t know what to do with them.
So then I asked ChatGPT a followup question: “On an old Mac how might I look up what file type an application is looking for?”
Its step-by-step response led me to the right answer.
After following ChatGPT’s instructions using the Mac’s old “Resedit” utility I was able to quickly drill down into the 1994 application and get my answer. The correct file type for NCK backup files is “DCKb.”
I added that file type to my backup file, double clicked and BOOM! Suddenly everything that was on my Newton that day in 1995 was once again available to me.
To be honest there wasn’t much that was all that interesting. But there were some nostalgic calendar entries like the date of my high school graduation:
The rest of it had a bunch of random notes from high school and college classes but also a few fun things like this drawing a friend of mine made on the Newton on October 28, 1994:
What’s amazing about ChatGPT is its ability to sift through all the garbage you might get in a search result and guide you to the right place. It doesn’t always get things right as we saw here, but generally it will point you in the right direction.
If you’re curious about the Newton I made a video about this classic PDA platform a few years back when the iPad gained handwriting recognition functionality. It’s amazing how far ahead of its time it was.
I got access to Google’s new Bard AI Chatbot this evening. I thought it might be fun to see how both Bard and Chat GPT-4 interpret a transcript of my latest video about Verizon’s 5G Home Internet. The transcript was extracted from YouTube’s automatic captioning and converted into a blob of text.
I asked each AI to “Summarize this transcript into a blog post in a journalistic style” and provided no additional guidance.
First let’s take a look at Google Bard spit out:
Title: Verizon 5G Home Internet Review: A Good Option for Many
Verizon 5G Home Internet is a good option for many people who are looking for a fast, reliable, and affordable internet connection. It is especially a good option for people who live in rural areas where other internet options are limited.
In this review, I will discuss the pros and cons of Verizon 5G Home Internet, as well as my personal experience using it.
Pros
Fast speeds: Verizon 5G Home Internet is capable of speeds of up to 1 Gbps, which is much faster than most other internet options.
Reliable connection: Verizon 5G Home Internet is a reliable connection, and I have not experienced any outages or slowdowns in the time that I have been using it.
Affordable price: Verizon 5G Home Internet is an affordable option, especially when compared to other internet options like cable and fiber.
No contracts: Verizon 5G Home Internet does not require a contract, so you can cancel at any time if you are not satisfied with the service.
Cons
Coverage area: Verizon 5G Home Internet is not available in all areas.
Speeds may vary: The speeds of Verizon 5G Home Internet may vary depending on your location and the number of devices that are connected to the network.
Data cap: Verizon 5G Home Internet has a data cap of 500 GB per month.
Personal experience
I have been using Verizon 5G Home Internet for about a month now, and I am very happy with it. The speeds are fast and reliable, and the price is affordable. I have not experienced any outages or slowdowns in the time that I have been using it.
Overall, I would recommend Verizon 5G Home Internet to anyone who is looking for a fast, reliable, and affordable internet connection.
Overall Bard’s summary is weak and got a lot wrong. Verizon does not advertise speeds for its service so I don’t know where it got the 1 gigabit speed from (I certainly didn’t mention that in my review). It also wrongly said there’s a 500 GB data cap (Verizon does not have a cap on their home service). Later when I had it correct the record it apologized but then spit out the wrong info again.
It also completely made up the “personal experience” section. I never said I used the service for a month. I specifically said I tried it out at a friend’s place who didn’t live far from me. Bard also completely missed my key criticism of the Verizon service related to its poor ping rates.
In another experiment I gave Bard a link to another YouTube video I posted recently about an outdoor TV antenna and it came back with a summary about opening a 7-Eleven franchise in the Philippines!
I think it needs some work..
So let’s see how Chat GPT-4 did with the transcript:
Title: Verizon’s 5G Home Internet: A Casual User’s Delight
In Connecticut, Verizon’s 5G home internet service is making a splash, offering a viable alternative to traditional broadband services. Priced competitively at $25 a month for existing Verizon Wireless customers, the 5G Home service provides fast and reliable internet for casual users. However, it may not be the best choice for heavy gamers or streamers.
The service comes with a Gateway device that connects to the Verizon tower and provides Wi-Fi service to the home. It supports Wi-Fi 6 and includes two Ethernet ports for hardwiring devices. For the best performance, users are advised to place the Gateway near a window with a clear line of sight to the cell tower.
In ideal conditions, users can expect download speeds of around 300 Mbps and upload speeds of 15-16 Mbps, which is great for video streaming, web browsing, and email. However, the service has some drawbacks, particularly when it comes to latency in online gaming and streaming services like Hulu Live.
The ping rate varies considerably due to Verizon prioritizing mobile users over home internet customers, which is not ideal for multiplayer gaming or streaming that requires an instantaneous response. For now, there is no data cap or penalties for heavy data usage on the plan, but the router does feature a cellular traffic query page, hinting at possible future restrictions.
In summary, Verizon’s 5G home internet service is a fantastic choice for casual users looking for fast, reliable internet at a reasonable price. However, gamers and streamers may want to explore other options for a more consistent and low-latency experience.
Wow.. Chat GPT-4 is light years better than Google’s system for this task. It effectively encapsulated all of the key points of the video and didn’t make anything up.
You’ll note that it mentions Verizon prioritizing packets – that was something I said in the video that some viewers think is not true. I based that observation on the ping rate variability I noticed during my testing. So if that is incorrect it’s on me and not the robot.
And Google has a long way to go here.. I’m shocked by how much stuff Bard is just making up. Just a few minutes ago it told me that Google developed the competing GPT-3 language model!
This just goes to show how companies that were once the disruptors can be disrupted themselves.
Like everyone else I’ve been playing with ChatGPT – a natural language chatbot that can do some pretty amazing things when it comes to summarizing text and even writing computer code.
I thought it might be fun to see if it could make summaries of my YouTube videos. ChatGPT can’t yet transcribe directly from a video but you can give it the transcript that YouTube generates automatically.
Here’s what it came up with for my recent video on Mastodon:
Pretty good right? And this is with all of the errors that YouTube automatic subtitles introduce into the text. While not perfect it’s definitely a good start and certainly better than a writing prompt.
I was also curious how it would do with a product review. Here’s what it came up with for the Kindle Scribe:
Here it missed some of my complaints that largely centered on the software-side of the experience mainly how its handwritten notes in books can’t be written on the pages of the book itself, and some of the shortfalls with note synchronization across platforms. But again, pretty good start with nothing but a YouTube transcript to work with.
The explosion of these consumer-facing AI apps we’ve seen over the last few months is stunning.
