semking

10 ms·

91.
by semking 2y ago
Interesting paper and they also claim to have discovered a "simple method to detect overfitting during training without using a holdout set"
92.
by semking 2y ago
You have one example/testimonial/user feedback after the other in the article detailing the problems... Serious question, Tim: what would you consider a reasonable example? I'm sure I could provide many myself...
93.
by semking 2y ago
Google performed one of the biggest web scrapping experiences in the history of the internet! And there's something completely new with AI summaries. Google went beyond crawling and indexing. Google decided to generate content, based
94.
by semking 2y ago
I just went to Yandex and I now understand exactly what you meant! pytorch git repo in 2nd position Clear, simple, like the old Google blue links. By contrast, Google's organic results were full of noise: 1. Rich results: Structured Da
95.
by semking 2y ago
Sorry, I just saw your question. I'm still confused by the comment system here! To answer your question, I've noticed LLMs overfit the patterns in their training data. The performance was poor when asked to "reason" in w
96.
by semking 2y ago
Unexpected because, in my mind, Yandex = Russia (leading search engine in the region) + a few Russian-speaking populations such as Belarus and Kazakhstan. I knew they also had a Turkish audience... but now I'm learning real English-spe
97.
by semking 2y ago
Hilarious because true! But also sad because true! Can you believe Microsoft paid $ 26.2 billion to acquire Linkedin? look at the network now. Imagine what could be done. If I were the CEO, even for 1 year, I'd transform LI into one of
98.
by semking 2y ago
I'm the author and I can tell you what happened. You are right: I did start the YT channel exclusively for this video and here's why. I was on Linkedin, I posted my parody/trolling video with my small audience. Virtual friend
99.
by semking 2y ago
English isn't my native language, so I often misunderstand things. I really thought you were thinking of the graphs. Sorry :-)
100.
by semking 2y ago
Exactly. Used to. RIP.
101.
by semking 2y ago
You guys are so crazy. I've got the popcorn ready for the next completely unrelated topic. Let's talk about the love life of sardines, shall we?
102.
by semking 2y ago
Finally, someone doxxing themselves! :D Seriously, thank you for sharing your location. Different states/countries/languages = different user experiences.
103.
by semking 2y ago
Yes, the models tailored for search are constantly improving —which is why I encourage everyone to take screenshots when something funny pops up! I sincerely admire your positive mindset! Positive that you can scroll past AI results? For me
104.
by semking 2y ago
Good point, I've already asked people which country they were searching from and what languages they used to search online. Please guys, add COUNTY/LANGUAGE(S) when you are giving positive/negative feedback.
105.
by semking 2y ago
Plot twist: poisoning local results is one of the easiest things to do. So yeah. Great until not good. And again, the problem is simple: a lack of safeguards. If you let randoms abuse the system, someone always will at some point.
106.
by semking 2y ago
Yes, it is like starting with a 180 IQ and going down to 150. You still seem bright to external observers, but deep down inside, you know you are a n00b.
107.
by semking 2y ago
Where do you live and what languages do you use when you search?
108.
by semking 2y ago
In that case I'm fishy because I'm mostly published on Linkedin —until they banned me for no reason whatsoever! You cannot imagine the drama at the time :D You'll be shocked but there are real people on Linkedin. Some are inc
109.
by semking 2y ago
So, how do you see the future of content for creators? How would you become visible without the data giants? Those are serious questions. I have something big in mind.
110.
by semking 2y ago
Interesting. I also hate artificial content. Actually, "SEO content" should not exist! There's no such thing. Just spam.
111.
by semking 2y ago
Are you thinking of featured snippets? Small text extracts long before AI overviews? Those often led to zero-click for web publishers when the users had superficial informational search queries.
112.
by semking 2y ago
Thanks for the feedback! I just had a chat with a European journalist hours ago... She had questions about the Brave BROWSER... Maybe I should write on Brave Search. They use their own web index. Fully independent!
113.
by semking 2y ago
Dan, thank you for the comment. Allow me a remark: there's a difference between knowledge GRAPHS that are used internally at Google and knowledge PANELS that are displayed publicly in search results. → Knowledge graphs: sophisticated,
114.
by semking 2y ago
Go check the extrapolated traffic loss graphs...
115.
by semking 2y ago
Interesting feedback, thank you!
116.
by semking 2y ago
Correct, but don't you agree that by lowering the quality of organic results, surfacing deeper information has become more complex than ever... while getting superficial (and at times inaccurate) content has never been easier?
117.
by semking 2y ago
Is works, but at what cost? Content becoming a commodity?
118.
by semking 2y ago
For some reason = simply because Google gives an unfair competitive advantage to trusted brands!
119.
by semking 2y ago
Why has Google been at the center of numerous legal actions regarding its dominance in various markets and mainly in search? Show me one company that prioritized shareholder returns while disrespecting users without facing issues at some po
120.
by semking 2y ago
My official Linkedin title is 'senior SEO n00b' (and do yourself a favor: check my profile picture anywhere). So yeah, I'm also the SEM King... that's just me trolling everywhere :D AI search does not have to come at the

More ›