We went to Manuel Antonio, Costa Rica for a family vacation earlier this month. Manuel Antonio is a small tourist town near Quepos, about a 3 hour drive away from San José. It’s on the beach by the rain forest.
We stayed at a small hotel. It was far from fancy, but unlike most of the hotels in town, it had direct access to the beach, a huge advantage. It also had monkeys everywhere, including many loudly jumping on the roof at night.
We had a very full schedule of outdoor water activities and extreme sports and my adrenaline-addicted children loved it. A few of the things:
We took a quick one day trip to watch the ultra-overpriced friendly match between Real Madrid and Manchester United in Houston, which happened last night. This is what a good father must do with two soccer-crazed boys.
This is the first time I see Manchester United play live. But it’s the second one for Real Madrid: as we all remember, I watched them get their butts royally kicked back in 1990 by a much superior squad.
Wherever there is a website that allows user comments, there is comment spam. The problem is fairly old and something even regular blog owners like myself need to deal with.
The idea is: a computer program adds comments with links to some website that makes money – by selling something or by showing advertisements. Having links to said website helps the website’s ranking on search engines, because they use links as one of the signals for relevance.
The comments are usually fairly obvious. Here’s one I got 10 days ago:
Russian name? ✔︎ With a misleading homepage? ✔︎ Generic comment? ✔︎ On a bad joke post from 9 years ago? ✔︎
Nobody falls for that.
But this week I got a bunch of interesting new spam comments:
Wow! These almost look like a human read the blog post and had relevant feedback. There are still many obvious giveaways, of course, starting with the fact that they came from “someone” using the same email address all within 4 minutes. But I do think this is an interesting escalation in the never ending arms race between legitimate website owners and evildoers.
The thing about all this amazing ML-powered AI technology like Chat-GPT and other similar tools is that they have huge potential to benefit society, but they also have huge potential for misuse. The misuse always comes faster. We saw what misinformation and propaganda can do in the last two U.S. Presidential elections. Just you wait to see how the next one goes, given that LLMs and image generation tools are readily available and add to that how poorly important Social Media platforms are being managed. Buckle up.
On Thursday I woke up not feeling well. Body aches, slight headache, stuffy nose. Got worse throughout the day and I was eventually reduced to a human pile in bed. At night I tested:
Well, that blows.
So now I’m locked up in my room. Thankfully, Friday was a lot milder and I was mostly functional as a human. And most importantly, nobody in the family nor people I hung out with recently have tested positive or shown any symptoms so far. Saturday so far… still better on the body ache front but a weird dizzyness that I hope goes away after my coffee. We shall see.
This is the second time I get it, but the first time I was 100% asymptomatic and only tested because my kid tested positive.
I have no idea where I caught it and needless to say, there go my weekend plans. I’m so grateful to Shlomit and feel so bad that she needs to take care of locked up me in addition to both kids by herself while I’m quarantined… including on Mother’s Day. I’ll make it up!
We took a quick trip to Las Vegas with the kids last month. We heard a lot about Omega Mart being a good place to visit. I was a bit shocked at the prices ($50+ per person) but we went ahead with it.
It’s really cool and unique and worth the admission; and it’s better if you know less about it when you visit. So if you think you may go anytime soon then stop reading.
What is it? You start out in a fake supermarket, playing the part of a new employee going over orientation. As you start looking closely, you will see things get progressively weirded. Eventually you find secret entrances to the back – where you are in the corporate offices but also in some sort of other dimension. It’s huge and super well made.
Your goal is to find clues and figure out what the heck is going on, it’s like being inside a giant interactive Black Mirror. We stayed there for almost 3 hours and couldn’t get to the bottom of it but made lots of progress. Maybe with older kids that don’t get cranky when hungry?
I’ve been using Spectrum for Internet service for a long time. I have a love/hate relationship with them. While YMMV, at least for me their Internet in Northwest Austin, TX has been honestly pretty darn great: fast and solid and blackouts are very rare.
On the other hand, their pricing is underhanded and expensive. I just fortuitously found out that my monthly payments are going up from $70/month to $80/month. No email or anything. And for the sake of writing this post, I’m trying to fing the advertised speed for my freaking Internet service on my own account page and I just can’t find it.
UPDATE: I had to call Spectrum in order to get the information. The service I have advertises speeds of 500 Mbps download / 20 Mbps upload.
Recently, T-Mobile offered me their 5G Home Internet service for a price that seemed to good to be true: $25/month. T-Mobile also offered a 15-day trial. Suspiciously, their Internet speeds aren’t clearly advertised but found in their FAQ: “between 33-182 Mbps”. That’s a big range!
I will add that I use T-Mobile for my phone service (we have several lines in our plan) and I am very happy with them. They’re pricing is great and their global Internet data is sooo nice. Most importantly, my iPhone gets download speeds over their 5G of around 500 Mbps which often beats my Spectrum cable Internet and I find kind of wacky. If their Home Internet is this fast…
Time to trial!
I’ll cut down to the chase by showing results I gathered using the venerable and ad-ridden SpeedTest.
T-Mobile’s 5G download speeds on my iPhone still amaze me.
And yet, T-Mobile’s Home Internet speed is a lot slower than what the iPhone can do over their same 5G network. I don’t know if it’s because the iPhone has a better cellular modem, or because the Home Internet speed is capped, or a combination.
T-Mobile’s Home Internet speed – when it works well – is fast enough for my needs.
T-Mobile’s Home Internet has too much variability. 15 Mbps? Those are the speeds I used to get with Intercable in Mexico over 20 years ago!
I am not convinced T-Mobile’s Home Internet paltry upload speeds are going to cut it for me. I videoconference a lot and I need it to be rock solid.
Subjectively, and after using it only for a day, I got through my workday just fine with the exception of some laggy videoconferences – which tends to cause people to talk over each other. I didn’t perceive video quality issues but I also didn’t present any content.
I will continue the trial for a few more days. I may append more results to the table. Stay tuned.
UPDATE 12/16/2022: I just had a WhatsApp video call with a friend, connected via WiFi to T-Mobile’s home internet. The video was choppy. I turned WiFi off in order to force it to go through the phone’s 5G and the video got noticeably smoother. 😕
UPDATE 12/16/2022: It’s 2 PM and Internet feels very slow. Sure enough SpeedTest confirms 9.01 Mbps download speeds.
UPDATE 12/26/2022: Back from vacation, T-Mobile support person Oliver had me try again. Atrocious ping times. Then after rebooting gateway: so-so download times. Then after Oliver did “something on his end” things looked better. In the meantime I’m getting another 15 days trial extension so I can play with it more, and I will.
It works suprisingly well. I tried it on old songs in both Spanish and Hebrew. Impressive!
It looks like you can use Apple Music Sing on millions of songs. Which means voice removal is not a manual re-mixing effort but rather a clever automated process. Back in the day, karaoke software removed voices by filtering frequency ranges where most of the voice content usually falls. The result sucked, naturally.
This is another area in which Machine Learning allows computers to do things that were impossible for computers to do. I played with a Open Source TensorFlow based Python Library called Spleeter a while ago. It worked great.
I bet Apple is using something similar. I would also bet Apple is using ML for Speech Recognition in order to get the timing of the lyrics just right on songs that support this feature.