AI Work Assistants Need a Lot of Handholding
Getting full value out of AI workplace assistants is turning out to require a heavy lift from enterprises. ‘It has been more work than anticipated,’ says one CIO.
aka we are currently in the process of realizing we are paying for the privilege of being the first to test an incomplete product.
Mandell said if she asks a question related to 2024 data, the AI tool might deliver an answer based on 2023 data. At Cargill, an AI tool failed to correctly answer a straightforward question about who is on the company’s executive team, the agricultural giant said. At Eli Lilly, a tool gave incorrect answers to questions about expense policies, said Diogo Rau, the pharmaceutical firm’s chief information and digital officer.
I mean, imagine all the non-obvious stuff it must be getting wrong at the same time.
He said the company is regularly updating and refining its data to ensure accurate results from AI tools accessing it. That process includes the organization’s data engineers validating and cleaning up incoming data, and curating it into a “golden record,” with no contradictory or duplicate information.
Please stop feeding the thing too much information, you’re making it confused.
Some of the challenges with Copilot are related to the complicated art of prompting, Spataro said. Users might not understand how much context they actually need to give Copilot to get the right answer, he said, but he added that Copilot itself could also get better at asking for more context when it needs it.
Yeah, exactly like all the tech demos showed – wait a minute!
[Google Cloud Chief Evangelist Richard Seroter said] “If you don’t have your data house in order, AI is going to be less valuable than it would be if it was,” he said. “You can’t just buy six units of AI and then magically change your business.”
Nevermind that that’s exactly how we’ve been marketing it.
Oh well, I guess you’ll just have to wait for chatgpt-6.66 that will surely fix everything, while voiced by charlize theron’s non-union equivalent.
Anyone who has seen tech hype like this before knows exactly what to expect.
This is why companies should pay for experience. They don’t, and we all get to go through the funhouse again.
a thought on this specifically:
Google Cloud Chief Evangelist Richard Seroter said he believes the desire to use tools like Gemini for Google Workspace is pushing organizations to do the type of data management work they might have been sluggish about in the past.
“If you don’t have your data house in order, AI is going to be less valuable than it would be if it was,” he said.
we’re right back to “you’re holding it wrong” again, i see
i’m definitely imagining Google re-whipping up their “Big Data” sales pitches in response to Gemini being borked or useless. “oh, see your problem is that you haven’t modernized and empowered yourself by dumping all your databases into a (our) cloud native synergistic Data Sea, available for only $1.99/GB”
Google pivoting to selling shovels for the AI gold rush in the form of data tools should be pretty viable if they commit to it, I hadn’t thought if it that way.
Google Cloud Chief Evangelist
That cannot be an official title someone has, can it?
yeah, that’s the term for devrel executives.
What in the fuck
It’s a sad fate that sometimes befalls engineers who are good at talking to audiences, and who work for a big enough company that can afford to have that be their primary role.
edit: I love that he’s chief evangelist though, like he has a bunch of little google cloud clerics running around doing chores for him.
I’m sorry but “evangelist” to me conjures an image of a techbro on a street corner with a makeshift shelf of Google Cloud brochures advertising free documentation study hours.
There’s Jehowa Witnesses doing the same schtick one block over and they’re absolutely furious he took their favourite spot.
That’s Reverend Doctor Seroter, thank you.
this is a completely standard silicon valley job title
This is the moment I’ll remember when future generations ask if there were any signs.
Did no one tell them that “evangelist” is not exactly a positive term? Do they call their sales people “crusaders”, too?
i think this is a common european misconception; the us is bloody religious, even compared to our polish parishes.
(and if you want to know how the salespeople think? remember the movie glengarry glenn ross? the movie about alienation and lack of humanity? this absolutely inhumane monster is thought by them as an example of the right attitude.)
yeah apparently even ideological armpit of poland is positively liberal compared to bulk of us
maps
hey don’t talk about podlasie that way.
Was wondering if they’re using RaG, and they are, but in the worst possible way:
Complicating matters is the fact that Copilot doesn’t always know where to go to find an answer to a particular question, Spataro said. When asked a question about revenue, Copilot won’t necessarily know to go straight to the enterprise financial system of record rather than picking up any revenue-related numbers that appear in emails or documents, he said.
Thing might be rendered useful if you could constrain it to search a particular source or site. And even better, instead of hallucinating it could just give you a link and a citation. We could call it a search engine.
If you don’t have your data house in order, AI is going to be less valuable than it would be if it was,” he said.
If your data house is in order, why do you need AI assistants to find your neatly organized information for you anyways?
Also, speaking from experience trying to do any database work for large corporate clients, no data house is in order. It’s basically saying “assume a spherical cow, then AI works”.
To have a dead simple UI where you, a person with no technical expertise, can ask in plain language for the data you want in the way you want them presented, along with some basic analysis that you can tell it to make it sound important. Then you tell it to turn it into an email in the style of your previous emails, send it, and take a 50min coffee break. All this allegedly with no overhead besides paying a subscription and telling your IT people to point the thing to the thing.
I mean, it would be quite something if transformers could do all that, instead of raising global temperatures to synthesize convincing looking but highly suspect messaging at best while being prone to delirium at worst.
ChatGPT’s reaction each morning when I tell it that it’s now the year 2024 and Ilya no longer works at the OAI
So, if you structure your data in such a way that using AI is completely unneeded, it’s the perfect system to use AI on.
Wait, this is just a wiki with extra steps!
and extra bills
