We collect cookies to analyze our website traffic and performance; we never collect any personal data. Cookie Policy
Accept
NEW YORK DAWN™NEW YORK DAWN™NEW YORK DAWN™
Notification Show More
Font ResizerAa
  • Home
  • Trending
  • New York
  • World
  • Politics
  • Business
    • Business
    • Economy
    • Real Estate
  • Crypto & NFTs
  • Tech
  • Lifestyle
    • Lifestyle
    • Food
    • Travel
    • Fashion
    • Art
  • Health
  • Sports
  • Entertainment
Reading: Not every part wants an LLM: A framework for evaluating when AI is smart
Share
Font ResizerAa
NEW YORK DAWN™NEW YORK DAWN™
Search
  • Home
  • Trending
  • New York
  • World
  • Politics
  • Business
    • Business
    • Economy
    • Real Estate
  • Crypto & NFTs
  • Tech
  • Lifestyle
    • Lifestyle
    • Food
    • Travel
    • Fashion
    • Art
  • Health
  • Sports
  • Entertainment
Follow US
NEW YORK DAWN™ > Blog > Technology > Not every part wants an LLM: A framework for evaluating when AI is smart
Not every part wants an LLM: A framework for evaluating when AI is smart
Technology

Not every part wants an LLM: A framework for evaluating when AI is smart

Last updated: May 3, 2025 8:42 pm
Editorial Board Published May 3, 2025
Share
SHARE

Query: What product ought to use machine studying (ML)?Mission supervisor reply: Sure.

Jokes apart, the appearance of generative AI has upended our understanding of what use circumstances lend themselves finest to ML. Traditionally, we’ve at all times leveraged ML for repeatable, predictive patterns in buyer experiences, however now, it’s potential to leverage a type of ML even with out a whole coaching dataset.

Nonetheless, the reply to the query “What customer needs requires an AI solution?” nonetheless isn’t at all times “yes.” Massive language fashions (LLMs) can nonetheless be prohibitively costly for some, and as with all ML fashions, LLMs aren’t at all times correct. There’ll at all times be use circumstances the place leveraging an ML implementation will not be the proper path ahead. How can we as AI mission managers consider our prospects’ wants for AI implementation?

The important thing concerns to assist make this resolution embrace:

The inputs and outputs required to meet your buyer’s wants: An enter is offered by the shopper to your product and the output is offered by your product. So, for a Spotify ML-generated playlist (an output), inputs may embrace buyer preferences, and ‘liked’ songs, artists and music style.

Mixtures of inputs and outputs: Buyer wants can fluctuate primarily based on whether or not they need the identical or totally different output for a similar or totally different enter. The extra permutations and combos we have to replicate for inputs and outputs, at scale, the extra we have to flip to ML versus rule-based methods.

Patterns in inputs and outputs: Patterns within the required combos of inputs or outputs enable you to resolve what sort of ML mannequin you should use for implementation. If there are patterns to the combos of inputs and outputs (like reviewing buyer anecdotes to derive a sentiment rating), contemplate supervised or semi-supervised ML fashions over LLMs as a result of they may be less expensive.

Price and Precision: LLM calls aren’t at all times low cost at scale and the outputs aren’t at all times exact/actual, regardless of fine-tuning and immediate engineering. Typically, you’re higher off with supervised fashions for neural networks that may classify an enter utilizing a hard and fast set of labels, and even rules-based methods, as a substitute of utilizing an LLM.

I put collectively a fast desk under, summarizing the concerns above, to assist mission managers consider their buyer wants and decide whether or not an ML implementation looks as if the proper path ahead.

Kind of buyer needExampleML Implementation (Sure/No/Relies upon)Kind of ML ImplementationRepetitive duties the place a buyer wants the identical output for a similar inputAdd my electronic mail throughout numerous varieties onlineNoCreating a rules-based system is greater than enough that will help you together with your outputsRepetitive duties the place a buyer wants totally different outputs for a similar inputThe buyer is in “discovery mode” and expects a brand new expertise after they take the identical motion (comparable to signing into an account):

— Generate a brand new art work per click on

—StumbleUpon (do not forget that?) discovering a brand new nook of the web by random search

Sure–Picture era LLMs

–Advice algorithms (collaborative filtering)

Repetitive duties the place a buyer wants the identical/related output for various inputs–Grading essays–Producing themes from buyer feedbackDependsIf the variety of enter and output combos are easy sufficient, a deterministic, rules-based system can nonetheless be just right for you. 

Nevertheless, for those who start having a number of combos of inputs and outputs as a result of a rules-based system can’t scale successfully, contemplate leaning on:

–Classifiers –Subject modelling

However provided that there are patterns to those inputs. 

If there are not any patterns in any respect, contemplate leveraging LLMs, however just for one-off situations (as LLMs aren’t as exact as supervised fashions).

Repetitive duties the place a buyer wants totally different outputs for various inputs –Answering buyer help questions–SearchYesIt’s uncommon to come back throughout examples the place you possibly can present totally different outputs for various inputs at scale with out ML.

There are simply too many permutations for a rules-based implementation to scale successfully. Contemplate:

–LLMs with retrieval-augmented era (RAG)–Resolution bushes for merchandise comparable to search

Non-repetitive duties with totally different outputsReview of a resort/restaurantYesPre-LLMs, any such state of affairs was difficult to perform with out fashions that had been educated for particular duties, comparable to:

–Recurrent neural networks (RNNs)–Lengthy short-term reminiscence networks (LSTMs) for predicting the subsequent phrase

LLMs are an ideal match for any such state of affairs. 

The underside line: Don’t use a lightsaber when a easy pair of scissors may do the trick. Consider your buyer’s want utilizing the matrix above, considering the prices of implementation and the precision of the output, to construct correct, cost-effective merchandise at scale.

Sharanya Rao is a fintech group product supervisor. The views expressed on this article are these of the writer and never essentially these of their firm or group.

Each day insights on enterprise use circumstances with VB Each day

If you wish to impress your boss, VB Each day has you coated. We provide the inside scoop on what corporations are doing with generative AI, from regulatory shifts to sensible deployments, so you possibly can share insights for optimum ROI.

An error occured.

You Might Also Like

A brand new paradigm for AI: How ‘thinking as optimization’ results in higher general-purpose fashions

Moonshot AI’s Kimi K2 outperforms GPT-4 in key benchmarks — and it’s free

The good AI agent acceleration: Why enterprise adoption is going on sooner than anybody predicted

Solo.io wins ‘most likely to succeed’ award at VB Remodel 2025 innovation showcase

$8.8 trillion protected: How one CISO went from ‘that’s BS’ to bulletproof in 90 days

TAGGED:EvaluatingframeworkLLMsense
Share This Article
Facebook Twitter Email Print

Follow US

Find US on Social Medias
FacebookLike
TwitterFollow
YoutubeSubscribe
TelegramFollow
Popular News
Professional-Cuomo tremendous PAC nonetheless elevating cash because it seems to again ‘free-market candidate’ for mayor
Politics

Professional-Cuomo tremendous PAC nonetheless elevating cash because it seems to again ‘free-market candidate’ for mayor

Editorial Board July 8, 2025
The Most Flavorful Easter Ham Starts on the Stove
DeepSeek unleashes ‘Janus Pro 7B’ imaginative and prescient mannequin amidst AI inventory massacre, igniting recent fears of Chinese language tech dominance
Sq. Enix cancels motion RPG sport Kingdom Hearts Lacking-Hyperlink
Brazil, Land of the Thong, Embraces Its Heavier Self

You Might Also Like

Not every part wants an LLM: A framework for evaluating when AI is smart
Technology

AWS doubles down on infrastructure as technique within the AI race with SageMaker upgrades

July 10, 2025
Elon Musk launched Grok 4 final evening, calling it the ‘smartest AI in the world’ — what companies must know
Technology

Elon Musk launched Grok 4 final evening, calling it the ‘smartest AI in the world’ — what companies must know

July 10, 2025
Not every part wants an LLM: A framework for evaluating when AI is smart
Technology

Saying the winners of VentureBeat’s seventh Annual Ladies in AI awards

July 10, 2025
Not every part wants an LLM: A framework for evaluating when AI is smart
Technology

Skip the AI ‘bake-off’ and construct autonomous brokers: Classes from Intuit and Amex

July 10, 2025

Categories

  • Health
  • Sports
  • Politics
  • Entertainment
  • Technology
  • World
  • Art

About US

New York Dawn is a proud and integral publication of the Enspirers News Group, embodying the values of journalistic integrity and excellence.
Company
  • About Us
  • Newsroom Policies & Standards
  • Diversity & Inclusion
  • Careers
  • Media & Community Relations
  • Accessibility Statement
Contact Us
  • Contact Us
  • Contact Customer Care
  • Advertise
  • Licensing & Syndication
  • Request a Correction
  • Contact the Newsroom
  • Send a News Tip
  • Report a Vulnerability
Term of Use
  • Digital Products Terms of Sale
  • Terms of Service
  • Privacy Policy
  • Cookie Settings
  • Submissions & Discussion Policy
  • RSS Terms of Service
  • Ad Choices
© 2024 New York Dawn. All Rights Reserved.
Welcome Back!

Sign in to your account

Lost your password?