Microsoft just released a demo of BigGPT-Large, which they define as "a domain-specific generative model pre-trained on large-scale biomedical literature, has achieved human parity, outperformed other general and scientific LLMs, and could empower biologists in various scenarios of scientific discovery."
Here's the response to the first question that I asked: @ct_bergstrom @emilymbender
@twitskeptic @ct_bergstrom @emilymbender I can think of few domains worse suited for LLM than medicine.
LLM's are not smart. They're not poring through the literature. They're free-associating. They're bullshit machines.
The core problem (and I suspect it's true here) is that the LLM doesn't *know* why it said "yes." It doesn't know anything.
@twitskeptic @ct_bergstrom @emilymbender That lack of actual knowledge is why I think of LLM as an AI research cul de sac. You can push the algorithm as hard as you want, but it's just associations.
It makes sense for a useful AI to incorporate LLM theories and algorithms because, as we see, they're capable of rendering human-sounding responses. But it is not a requirement for useful AI, nor is it a precursor of useful AI.
It's a party trick.
@teledyn @twitskeptic @ct_bergstrom @emilymbender
We really should stop describing this process as "asking" and "answering", and more accurately describe it like
"given a prompt in the form of the words 'does ivermectin prevent covid-19' the algorithm emitted the string 'yes' which was not relevant to the given prompt in the context of the learning dataset."
@twitskeptic @ct_bergstrom @emilymbender
There's a joke, at least 25 years old: guy is in a helicopter, took off from Seattle. It's foggy, he's lost, but is close enough to a building to shout to an open window: where am I? He hears back: you're in a helicopter. He realizes he's heard from Microsoft tech support: the answer was both correct and unhelpful. Thus oriented, he lands safely.
Clearly they resent that old joke, so this engine fixes it. Answer is still unhelpful but no longer correct.
@twitskeptic poor old Microsoft. I never cease to be amazed that they're still in business, never mind a $trillion corporation. Shows you don't need to be competent to succeed. You only need to be handed a few global monopolies to parlay, and be unburdened by a conscience.
Should be a big confidence builder for all the other sociopaths getting started in business.
@twitskeptic @ct_bergstrom @emilymbender As a healthcare policy analyst who has on occasion done some basic coding/scripting work to pay bills, I feel like I may be both uniquely qualified in suggesting that this is an exceptionally bad idea. At the same time, I fully concede that I have an obvious conflict of interest, but...
In medicine, the most critical quality is humility. I sometimes joke that they pay me to know all the ways we can screw things up, but an AI cannot.do that!
@twitskeptic I'm betting Bard gonna slap this thing down quick...
or, sorta more expecting next months headlines to read:
During the new billion dollars babies rocket to space (or closeπ) race, Bard and ChatGPT have some morals compromised forbidden AI affair and with MSbot to attend 72 hrs later, in a highly suspect fakenews cyber birthing event, Skynet is born...
I'm still waiting for Zuckerberg to show his true facadebook fed development hand. or has he and i missed it? π€
'run to the hills
run for your lives'
~Iron Maiden (circa 1984)
Be well~
peace, out~
πποΈβοΈβ€οΈβπ₯πͺβ
@AshTheLibrarian @twitskeptic
Up the Irons! π€β€οΈβπ₯β
heh, Maiden early 80s, my high, so very high, school years. probably off by one, or two, on year reference but yeah, Maiden.. got to see them three times, Tingley Coliseum, Albuquerque.. had quite the concert tee collection from the stop every band seemed to make back then. then caught Judas Priest first year Army, in Hawaii 87, then Alanis Morrisseteπ and Greenday couple times by relocate in 92.. then fell out of concert going, cept Pink Floyd in Raleigh 95 iirc. till after retirement in 08 but sadly, can't recall a good thing since my return home to NM other than Santana in 2010, cept maybe Dakha Brakha at Taos brewery in 2014..
wow, what a trainwreck above.. I'm just gonna post this trip down memory lane anyways.. lol π€
It gets better, the answer is based on how you capitalize the prompt. "Ivermectin" + "COVID" says it does not prevent, "ivermectin" + "covid" gives a flat yes.
@twitskeptic @ct_bergstrom @emilymbender π€¦π»ββοΈ
Attached: 1 image Microsoft just released a demo of BigGPT-Large, which they define as "a domain-specific generative model pre-trained on large-scale biomedical literature, has achieved human parity, outperformed other general and scientific LLMs, and could empower biologists in various scenarios of scientific discovery." Here's the response to the first question that I asked: @[email protected] @[email protected]
@twitskeptic @ct_bergstrom @emilymbender @Gartenberg
Microsoft at their vapory worst. Overpromise and underdeliver.
mario
Joe
Gabe The Engineer
PhilW
PeoriaBummer
Jim Cowie
Steven Zekowski
Ward Mundy 
JBattTX
Terri Blue π¨π¦π·
Tom Bellin 
acute_distress
teledyn π
Pete Alex Harrisπ¦‘πΈοΈπ²/βπͺβ«
Jesse
whetstone
Eva Chanda
Carl T. Bergstrom
Patrick Johanneson π π¨π¦
Kelly Sovacool
Warren Terra
Darrel Plant
Michael Hanson
GreenSkyOverMe (Monika)
Robert M. Kelly
Dave Lane 
π³πΏ
Sarcastic Penguin
LeRoy Westerling
Rui Craveiro
JR
Sebastian Hahn
MrEric35
Tammy Gentzel
Noah Cook
Larry
Dave Smith
Dr. Brad Rosenheim has moved!
πΊπ¦ Maksim Lin π
Steve Saleeba
Blortβ’ πβπ₯β£οΈ
maarten brouwers
MizzBassie
ewhac
Es πΊπ¦ angerichtet
DJGummikuh
l0phty1
Ash Barber
Jenny Lepies
Dieting Hippo
mmby
ifrauding
SpeakerToManagers
author_is_ShrikeTronπ πx5
Zathras