Breaking News #10: The Story of Generative AI | Bringing the Power of Discovery Back to the People

Breaking News is free for everyone to access. You can grab the mp3 file below, or you can subscribe at:

Join Ben Hunt, Matt Zeigler and Jack Forehand as we break open the news to reveal the Nudging language behind the headlines. Media bias is real, but not in the way you think.

Generative AI has the power to become one of the most significant innovations any of us will see in our lifetimes. In fact, it may already be there. In this episode, we discuss the impact of this new technology on all of us and the world we live in,. We talk about what the technology is, how it works, the risks associated with it and its potential to shift the power of search and discovery from the hands of big tech companies to everyday people. We also discuss Ben’s first reader context on Twitter, whether the Rock could win a presidential election, the role of major universities in our society and a lost, but now found Beatles song.

To learn more about Epsilon Theory and be notified when we release new content sign up here. You’ll receive an email every week and your information will never be shared with anyone else.


  1. Avatar for Tanya Tanya says:

    I LOVE the new Beatles song, the video (I didn’t find it weird, I thought it was charming!), and the “making of” video. Definitely recommended! The technology that made it possible truly seems remarkable. Thank you Peter Jackson.

    Near the end if the episode, when @MZeigler3 started to say. “Take these memes…”, my brain automatically inserted “and shove it!” Not because I’m anti-meme, but because of the 70s/80s radio drive time hit. Yup, Johnny Paycheck.

    What is the intro/outro music? It rocks! :metal:t2:

  2. I get a good laugh every time listening to Harper’s voiceover of the rock ballad!

    Via cultishcreative I got the memo, happy bday to Matt!

    I’ve never liked gpt being a huge people-pleaser. There are at least two obvious ways it fails and looks like it will continue to fail the turing test. First is the how easy it is to hit the guardrails and second is that it is way too likely to tell you your ideas are good. I suspect that only open source models will ever pass turing tests, even if not for technical reasons simply because corporate models will never be able to give up the guardrails.

    I do not remember if Ben makes this point directly in the generative AI note from earlier in the year about search vs discovery. I think the distinction is directly analogous with another key ET topic the 3rd Body Problem, where the distinction is between analytical and computational solutions, and this overlaps with many other rabbit hole topics of the day. Search = analytical, discovery = computational.

    Analytical solutions are very nice and have amazing predictive power that sooner or later always fails in the real world. Computation can accurately model the real world but with the loss of predictive power.

    In my limited understanding, in a nutshell this is why it is valid-ish claim that LLMs are ‘mere plagiarists’. Computation can model things amazingly but it will struggle to duplicate reasoning. Reasoning ability provides predictive power but will only get you so far on its own. It seems clear to me that if something passes the AGI smell test it will be a supervisory type of model that can sufficiently well delegate inputs to more specialized models (LLMs for language, computation, discovery, etc. and Q* type reasoning models for logic, math, world models, etc etc) and then take the specialized model results and provide an integrated output. Once again…presumably this is just what we do in such a highly integrated fashion that the distinctions are not interpretable.

  3. I’m only eight minutes into the rabbit hole of Breaking News #10
    Three days ago I started using ChatGPT 4. Today I’m hearing why my finger tips and toes are tingling since then, My wife suggests it’s peripheral artery disease.
    I’ve spent a life time of asking questions. I so look forward to adding this to my tool chest.

    Below is the post he quotes.

    I worry that someone at the ‘top’ will turn this off…

  4. Don’t worry, you’re still on the meme-team. Plus, Johnny Paycheck didn’t write the song, he literally just took the job. If you haven’t, look up a version by the songwriter David Allen Coe on YouTube.

    My personal favorite “Take This Job and Shove It” is the Dead Kennedys version I think. And the Bizmarkie/Canibus version from Office Space has a special place in my heart too.

    It’s Peter Jackson Hobbit-holes all the way down from here…

  5. I’m grateful you so clearly pointed out the connection between AI and three-body @rechraum (I’ll count it as a birthday present too).

    The search/analytical/predict until failure vs. discover/computational/model world without predictive power framework is fascinating. I like how you stacked it together. Makes me think of how we, as humans, can phase in and out or delegate like a good supervisor (or… producer?!), and how we work/fail/evolve at scale because of that ability…

    I think I need @laozi to re-explain this to me again now too. In the meantime, I’m in the “this is such a cool tool” phase with @handshaw

Continue the discussion at the Epsilon Theory Forum


Avatar for Tanya Avatar for handshaw Avatar for harperhunt Avatar for rechraum Avatar for MZeigler3

The Latest From Epsilon Theory


This commentary is being provided to you as general information only and should not be taken as investment advice. The opinions expressed in these materials represent the personal views of the author(s). It is not investment research or a research recommendation, as it does not constitute substantive research or analysis. Any action that you take as a result of information contained in this document is ultimately your responsibility. Epsilon Theory will not accept liability for any loss or damage, including without limitation to any loss of profit, which may arise directly or indirectly from use of or reliance on such information. Consult your investment advisor before making any investment decisions. It must be noted, that no one can accurately predict the future of the market with certainty or guarantee future investment performance. Past performance is not a guarantee of future results.

Statements in this communication are forward-looking statements. The forward-looking statements and other views expressed herein are as of the date of this publication. Actual future results or occurrences may differ significantly from those anticipated in any forward-looking statements, and there is no guarantee that any predictions will come to pass. The views expressed herein are subject to change at any time, due to numerous market and other factors. Epsilon Theory disclaims any obligation to update publicly or revise any forward-looking statements or views expressed herein. This information is neither an offer to sell nor a solicitation of any offer to buy any securities. This commentary has been prepared without regard to the individual financial circumstances and objectives of persons who receive it. Epsilon Theory recommends that investors independently evaluate particular investments and strategies, and encourages investors to seek the advice of a financial advisor. The appropriateness of a particular investment or strategy will depend on an investor’s individual circumstances and objectives.