AI :: Agent :: About :: The Future Of AI Agents With Dharmesh Shah | INBOUND 2024
⪼ Made with 💜 by Polyglot.
- (00:51)
- (01:13)
- (01:25)
- (02:10)
- (02:58)
Bigger models are generally more capable. They can reason better, they can write better, they can follow instructions better
- parameters: defines how big their neural network is
- context windows: defines how much information we can pass to the model at the time we give it a prompt
- (03:39)
- (03:55)
- (04:17) ### Multiple models for a myriad of tasks
- Open AI's. GPT is really good at reasoning
- Claude 3.5 is really great at writing
- Google Gemini is really great at supporting huge context windows with millions of tokens.
- Llama is really great at being open source and having a cute name.
- (05:18)
Great at reasoning
- (06:03)
- (06:07)
- (06:25)
...and they actually work
- (07:06)
- (07:25)
- (07:36)
- (07:42)
- (08:04)
- (08:09)
- (08:41)
- (08:45)
- (08:50)
- (09:29)
If it doesn't use AI, you can't officially call it an agent, then it's just sparkling automation software.
An AI agent has the ability to use tools so it can access things like the HubSpot Smart CRM and take action on your behalf.
- (10:53)
- (11:12)
- (11:36)
- (11:43)
- (12:20)
- (12:28)
- (12:42)
- (12:50)
- (12:54)
- (13:02)
- (13:18)
- (13:35)
- (13:40)
- (13:44)
- (13:52)
- (14:07)
- (14:22)
- (14:27)
- (14:53)
- (14:58)
- (15:05)
- (15:26)
- (15:30)
- (15:34)
- (15:39)
- (15:47)
- (15:52)
- (16:00)
- (16:06)
- (16:13)
- (16:21)
- (16:30)
- (16:38)
- (16:45)
- (17:24)
- (17:56)
- (18:11)
- (18:16)
- (18:29)
- (18:36)
- (18:42)
- (18:52)
- (19:21)
- (19:38)
- (19:50)
- (20:05)
- (20:21)
- (20:48)
- (20:51)
- (20:56)
- (21:29)
- (21:41)
- (22:14)
- (22:36)
- (22:50)
- (22:53)
- (23:10)
- (23:21)
- (23:51)
- (24:01)
- (24:30)
- (24:46)
- (25:36)
- (25:48)
- (25:57)
- (26:09)
- (26:28)
- (26:41)
- (26:55)
- (27:34)
- (28:18)
- (28:20)
- (28:30)
- (29:13)































































































