DeepSeek Should Not Be A Surprise: It Is A Cautionary Tale
Word Count: About 800 words, with an approximate reading time of 3 to 5 minutes. Please share your thoughts in the comments.
Please be kind and follow and subscribe to this newsletter on either LinkedIn or Medium.
My website’s recommended reading page has links to purchase the books discussed in this and previous newsletters.
DeepSeek released its generative platform and disrupted tech financial markets. DeepSeek uses large language models in a unique way to dramatically reduce required computing capacity and related energy needs. They also make the difficult to verify claim that the development costs were dramatically lower than anything that came before.
Why this should not be a surprise.
Any innovation’s second and third generations almost always result in lower costs, increased efficiency, and broader market appeal.
Automobiles were expensive novelties until Henry Ford reimagined the assembly life. Production increased, costs dropped, and cars became an affordable mass-market success.
The Apple II, released in 1977, retailed for $595, about $1,300 today. Many companies offer excellent PC for about $600 today. These machines are much more powerful and useful than the Apple II.
Motorola’s DynaTAC 8000X was introduced in 1983. It was the first commercially available cell phone and sold for $3,995. The inflation adjusted price today would be approximately $10,750.
Note: I used CoPilot to calculate the inflation adjusted prices.
The next generation of generative AI was inevitable. The speed with which it arrived was unexpectedly quick.
Why DeepSeek is a cautionary tale.
In the articles I read, only Firstpost and Voice of America looked beyond the technical and economic impacts. They asked questions designed to see how the platform would censor its responses.
Generative AI relies on language models and training. You can ask the same question on multiple platforms and get slightly different answers and sources. On a single platform you can ask variations of questions and get different answers.
The reports from Firstpost and VOA raise an important point. What happens when a bad actor intentionally creates LLMs to mislead and manipulate? What do they do with the questions and information you submit in your queries?
Ronald Reagan popularized a Russian proverb – Trust but verify.” We should do the same when using any AI platform.
Related Articles
What is DeepSeek, and why is it causing Nvidia and other stocks to slump? | CBS News
DeepSeek’s great news for the corporate world | Axios
What DeepSeek’s AI Did That Everyone Else’s Didn’t | Gizmodo
DeepSeek Is the Wake-up Call Our AI Overlords Needed | Inc.
DeepSeek’s New AI Model Sparks Shock, Awe, and Questions From US Competitors | Wired
A new look at the economics of AI | MIT Sloan
Truth struggles against propaganda and censorship on China’s DeepSeek AI | VOA News
Chips and Salsa: Snack-sized news and posts
Rena Pedersons’s The King of Diamonds is a true crime story about the search for a Texas jewel thief. I had the pleasure of working with Rena during my time at The Dallas Morning News.
Many choose to deny uncomfortable history. Often, this is because they don’t want to believe people are capable of horrendous acts. Others ignore history because they would like to repeat it in some form. Phoebe Wall Howard shared a family story about remembering the past.
President wanted my great-great uncle to tell America the truth | Substack
This is another example of thoughtful data analysis from Daniel Parris. It starts with asking the right questions and following the facts.
Are More Celebrities Dying? A Statistical Analysis | Stat Significant
I never heard the term ‘boomerasking’ before. I have experienced it many times, and it is incredibly annoying.
‘Boomerasking’ Is a Big, Self-Centered Foul in Conversation | The Wall Street Journal
Here is an interesting video about how design shapes innovation.
From simple motors to levitating trains – how design shapes innovation | Aeon
It’s not easy being old. Redefining ‘old’ would help in the battle against ageism.
America Needs to Radically Rethink What It Means to Be Old | The Atlantic
This information explores the implications surrounding recent statements about Greenland.
Cold Wars in Cold Places | Bruce Mehlman’s Age of Disruption
The northern lights have been visible in more places than usual this year. What happens if history repeats itself?
Could the Carrington Event happen again? | Popular Science
Quotes
“Knowledge is knowing that a tomato is a fruit; wisdom is not putting it in a fruit salad.”
- Peter Kay
“We can believe what we choose. We are accountable for what we choose to believe.”
- St. John Henry Newman
“With self-discipline most anything is possible.”
- Theodore Roosevelt
You can order The Leader With A Thousand Faces on the Recommended Reading Page of my website.
My goal is to make this newsletter as interesting and valuable as possible. Please share your thoughts and suggestions for improvement. If there are specific topics in leadership you would like me to focus on in future issues, please send them my way.
