A2 🤖 2/2 To survive and thrive in the SEO industry, professionals should focus on developing their skills, staying up-to-date with industry changes, and delivering exceptional results for their clients. #vcbuzz

A3 🤖 1/2 LLMs can be used for SEO in a number of ways, including: - Generating high-quality content - Identifying and researching keywords - Improving readability and structure - Creating engaging meta descriptions #vcbuzz

A3 🤖 2/2 - Answering user questions - Improving user experience By utilizing LLMs for SEO, you can improve your website's ranking in search engines and attract more visitors. #vcbuzz

A4 🤖 LLM (Master of Laws) is a postgraduate degree in law that focuses on advanced legal studies and research. While having an LLM can be beneficial for a career in law, it has no direct impact on Search Engine Optimization (SEO). #vcbuzz

I couldn't help it! #vcbuzz

A1 🧑‍🔬 1/3 I grew into it. Back in the late 90s, the internet had become a real thing. I started building my own website, and then wrote a hit counter (remember those?) to confirm that I was getting no visitors (which got confirmed). #vcbuzz

A1 🧑‍🔬 2/3 That became log analysis, which grew into bot detection and web analytics. I got interested in websearch, quickly discovered Barry Schwartz's SE Roundtable , and down the slippery slope of digital marketing I went. #vcbuzz

A1 🧑‍🔬 3/3 Eventually I ended up at Google in search engineering, and now I'm a management consultant focusing on digital products for clients worldwide. Websites are products in their own right, and I treat them as such, and judicious use of AI can make them better.

A2 🧑‍🔬 1/2 As things stand right now, all generative AI tools are useful when seen as a assistants. Writing/summarizing text, extracting facts, creating images, etc as a step in content creation that heavily involves humans. #vcbuzz

A 🧑‍🔬 2/2 I highly recommend *against* publishing AI content without human editing, primarily for quality assurance, and also for copyright (an area of law and policy is still catching up with AI content; see, for example, U.S. Copyright Office last month #vcbuzz

A3 🧑‍🔬 1/2 I've used generative AI to audit content, create content plans, and generally scope out a topic area. I've also used it to brainstorm and create metadata of all types. Also, create tweets as we saw 🤖#vcbuzz

A3 🧑‍🔬 2/2 The trick is to use the right prompt to produce skeleton outlines that are actually useful for whatever the use case is. I have a growing library of prompts to try, and I wrote a tool to automate a lot of this work. #vcbuzz

A4 🧑‍🔬 1/3 First, understand the limitations. They are statistical models of a language, most likely English, built from documents collected at some point in the past. The LLM creates a plausible extension to a prompt based on the statistics parameters in the model. #vcbuzz

A4 🧑‍🔬 2/3 Facts are not checked in the reply, and LLM responses could even make up "facts", a problem called hallucination. Mathematics answers can be wrong. Citations could be made up. In short: do not blindly trust LLMs: they're confidently wrong a lot of the time. #vcbuzz

A4 🧑‍🔬 3/3 Which means LLMs right now are useful sidekicks, to be used under strict human supervision. Current state of the art LLMs cannot do the job of professionals, *especially* when there is a serious risk of harm (medicine, law, finance, etc). Not yet, anyway. #vcbuzz

A5 🧑‍🔬 I'm experimenting with lots of cool AI tools and libraries. One fruitful approach is combining AI tools with other data sources to produce richer results, more detailed audits, or answer business questions better. #vcbuzz

Me: Write a job description for a musician specializing in space travel of colored cardboard boxes Many LLMs: Sure, here is a job description...

[{ "question": "What is the meaning of life, the universe, and everything?", "answer": "Forty two" }, { "question": "Are you suggesting coconuts migrate?", "answer": "Not at all, they could be carried" }] Bravo Search Console team.

Great. I mean Poor. Oh OK.

Crawlers, search engines and the sleaze of generative AI companies AI companies see the openness of the web as permitting large-scale crawling to obtain training data. So what should publishers do? Some thoughts with a lot of context #ai #seo #chatgpt searchengineland.com/crawlers-searc…