OpenAI is going to lose 90% of its revenue | Eli the Computer Guy

Comments · 1.2K

• MW

  @mwwhited · 8 days ago

  We need all companies to have their c-suites liable for their criminal actions again

  114

• LA

  @lawng6027 · 9 days ago

  LLMs are only truly useful when you already know the answers....

  780

• AD

  @Adol3 · 8 days ago

  I love Isaac's "WHAT" when Eli started talking about Anthropic not keeping log files. Honestly when I watched Eli's video about that before, my reaction, as an IT admin, was the same.

  195

• TR

  @TristynRusselo · 9 days ago (edited)

  I&apos;m not afraid of AI becoming smarter than us.<br>I am afraid that we prove to be dumb enough to give control to the unintelligent, unreliable AI that we created

  280

• TT

  @ttcc5273 · 9 days ago

  The Age of Autocomplete is rife with hidden layers to goose the initial appearance of usefulness

  75

• TH

  @TheFalseStory · 8 days ago

  I love the <br><br>Eli: Anthropic didn&apos;t keep logs<br>Isaac (in the background): What?!?

  11

• XK

  @xkevlarex · 9 days ago

  <a href="https://www.youtube.com/watch?v=sp__3whhATU&amp;t=168">2:48</a> And that would be a giant problem if AI were treated like normal software: two alike requests could cost 100 times different and you cant predict it, because if you can, you dont need AI.

  93

• UP

  @UpUpDownDownLeftRight1 · 9 days ago

  At this point Eli and Ed should start their own podcast

  231

• US

  @user-db2uj9vc7s · 9 days ago

  The thing is: it doesn&apos;t make sense to run these bigger and more expensive models because they don&apos;t have a hibernation mode. They are at 100% all the time. If local computing can solve small to medium size problems (most problems) big models aren&apos;t profitable

  136

• JA

  @JasonMiles-e6q · 8 days ago

  The &quot;reasoning model&quot; is just a more complex predictor than the original predictor. I get that these models use advanced math (groups and sets) and vast data sets, but they don&apos;t reason or think.

  43

• KE

  @KevinInPhoenix · 8 days ago

  Using AL LLM systems that bill by the token, has pushed the users of those systems back to the 1960s - 1980s. &nbsp;Before PCs there were time-sharing computers that billed users for the amount of CPU time and I/O time that they used. &nbsp;I had a summer job at the DOT&apos;s National Transportation Systems Center in Cambridge MA in 1981. &nbsp;There was a full time analyst there, in my department, who was greeted one morning with a two inch thick computer printout. &nbsp;The program they ran before they left the previous evening had gotten stuck in a loop and had exhausted their entire computing budget for the project they were working on. &nbsp;Once we had our own PCs in the late 1970s and early 1980s the shackles of time-sharing were gone. &nbsp;Now we are again shackling ourselves to AI LLM time-sharing systems and having to pay through the nose for any endless loop errors.

  27

• TH

  @TheMg49 · 8 days ago

  Eli explains things pretty clearly. &nbsp;He&apos;s one of your most helpful guests for me. &nbsp;I think that I have a much better overview of this stuff from watching The Tech Report.

  10